An analysis of savings and investment of the farming household in Bangladesh. Draft Report (Please do not quote)

An analysis of savings and investment of the farming household in Bangladesh Draft Report (Please do not quote) Report Prepared by Pollibir Unnayan Sangstha, Thakurgaon, Bangladesh (Pollibir Development Organizations) for the Food and Agriculture Organization of the United Nations (FAO) Rome, Italy, February 2013

Table of Contents Main findings 3 1. Introduction 4 2. Agriculture and food security in Bangladesh 6 2.1 Macro economic scenario 2.2 Sectoral overview of agriculture 6 2.3 Food security situation in Bangladesh 14 Page 3. Investment in agriculture 17 3.1 Overall Investment in agriculture 18 3.2 Public investment in agriculture 4. Analysis of Savings of the farming households 19 4.1 Ownership of land 19 4.2 Distribution of household by occupation 19 4.3 Income, expenditure and savings by size of own land in rural area 20 5. Saving behaviour of households with and without migrant workers 21 6. Household income and expenditure based household income 23 and expenditure survey 6.1 Level of income by household sizes 24 6.2 Sources of income 27 6.3 Consumption expenditure 28 6.4 Food expenditure 30 Annexes 33 A1. Sample design 33 A2. Concepts and definitions 34 2

Main findings If income less expenditure is counted as saving then, overall, based on a nationwide survey, it appears that in Bangladesh the farming household in aggregate do not save. Household income and expenditure survey shows that for all group of households, the percentage of income saved is almost nil. Household with no land and those with land but with holding size of less than half a acre, nearly 66 percent of the holdings of the country, have negative savings. The holding size of up to 1.5 acre also do not have any positive savings. Therefore if this group is also included then it would appear that all households with less than 1.5 acre of land, amounting to nearly 85 percent of the farming households, including the landless, do not have any positive savings. Only the households with holdings of 1.5 acre and above generate positive savings, and higher the holding size higher the rate of savings. What emerged is that in terms of savings and investment, in addition to holding sizes, there is a clear distinction between the families with a migrant and those without a migrant. Average annual household saving for migrants amounts to Tk.87583 compared to Tk. 28957 for non-migrant households. Thus migrant households save about 25 percent of their total annual income while non-migrant households save about 22 percent of their total annual income. There is large difference between absolute saving of these two groups of households; the difference between the proportions of income saved is small. Savings are held in different forms, the most popular form being bank deposits. Agricultural households both migrant and non-migrant, invest in agriculture as well as outside agriculture. Agricultural households hold real as well as financial assets. They also reduce their financial obligation. As expected the migrant households spend a large amount on land. It was pointed out during FGD that migrant households are not limited to purchase of land in their own village or locality; they also buy land in urban areas. Acquisition of assets by migrant households is significantly different (higher) from that of non-migrant households in most cases. 3

1. INTRODUCTION According to 6 th Five Year Plan of Bangladesh (SFYP) for 2011-1015, one of the major problems Bangladesh economy is facing today is the stagnation of the overall level of investment in the domestic economy (Figures 1 and 2). Aggregate investment has stagnated in the 24%-25% of GDP range in recent years, despite a steady increase in the national savings rate. Although private sector investment has been increasing at a pace slightly above the rate of growth of GDP, a secular decline in public investment in relation to GDP largely offsets that, keeping total investment broadly stagnant in relation to GDP. This low level of investment significantly falls short of the investment rate needed to support the 8% GDP growth target set for in the 6 th Plan. Figure 1: Pattern of Private Investment Growth Source: Bangladesh Bureau of Statistics Figure 2: Public Investment and ADP in relation to GDP Source: Ministry of Finance Achieving the higher growth target will require total investment under the Plan to increase steadily by 8.1 percentage points in relation to GDP to 32.5% of GDP by FY15. In addition to the private sector, the public sector will play a catalytic role in raising the total investment rate to the required levels during the SFYP period. Public investments and policies would create the necessary investment climate and heighten investors (both national and foreign) confidence to undertake the required investments. 4

As in the past, much of the additional increase in the growth of investment is projected to come from the private sector. The secular increase in the relative share of private investment in total investment is in part a reflection of reforms initiated in late 1980s and in 1990s by removing restrictions, initiating privatization of public enterprises, and creating a more favourable investment climate. The growing share of private sector investment reflects favourable private sector response to the improved investment climate. However, the rate of growth of private sector growth in investment has slowed down in recent years due to infrastructure constraints which tended to intensify on account of declining public investment and the inability of the public sector to undertake large infrastructure projects (The 6 th Plan). The declining trend in public investment in relation to GDP is a matter of concern. Past difficulties in ADP implementation (Table 1) prevented the government from investing in many critical areas such as infrastructure and agriculture. Over the years the infrastructure gap has been widening and has become a binding constraint by choking Bangladesh s economic growth potential. Table 1: Budget ADP Allocation and Actual Spending, FY05-FY11 % Increase over the Fiscal Year Budget Actual Actual as % of Budget previous year (In billion taka) (%) (%) FY05 227 194 85.3 1.8 FY06 246 175 71.0-9.6 FY07 260 176 67.5 0.5 FY08 250 184 73.8 4.8 FY09 256 193.7 75.7. 5.3 FY10 305 255.4 83.7 31.8 FY11(estimate) 385 358.3 93.1 40.2. Source: Ministry of Finance Efforts to reverse the declining trend in ADP was launched with the FY10 budget and intensified further in FY11 budget, the first year of SFYP. The Government established an ambitious target for ADP in FY10 by setting the ADP target at Tk. 305 billion, which was about 57% higher than the outturn of the preceding year. Although the actual increase was lower (31.8%), this was a major improvement over the ADP implementation rate recorded in recent years. After many years of steady decline, ADP spending in relation to GDP increased by 0.5% in FY10. Building on this gain, the target for ADP in FY11 has been set at 0.8 percentage points higher at 4.9% of GDP. The Plan s objective is to sustain this momentum by increasing the ADP size by an additional 2.4 percentage points to 6.1 % of GDP. Special emphasis has also been given to infrastructure and power sectors in the FY11 ADP. Implementing this larger ADP, while maintaining the quality of spending, will certainly be a major challenge. To ensure the quality of spending and better project implementation, all ministries are being brought under the Medium Term Budget Framework (MTBF). The planning and budgeting processes are being strengthened to improve the quality of public investment. Aggregate Savings The sharp rise in investment projected in the SFYP will be largely supported by a significant increase in national savings. National savings, comprising domestic savings and inflow of workers remittances, have been on a rising trend owing to increased domestic saving but also because of rapid growth in the inflow of remittances (Figure 3). As the national savings rate increased the domestic investment effort 5

did not expand commensurately causing the external current account balance to reach a record surplus of 3.7% of GDP in FY10. Figure 3: Gross National Saving Rebound and its Key Drivers Source: Bangladesh Bureau of Statistics and the Bangladesh Bank Building on the recent positive performance on the national savings front, the Plan aims to increase national savings rate by 2.1 percentage points to 32.1% of GDP. The pace of increase is about the same as recorded in recent years. The fall in the savings rate in FY11 was primarily attributable to the slowdown in remittances and a marked increase in imports associated with the global commodity price increase. The savings rate should revert back to the level of the base year by FY13 as global prices stabilize and growth in the inflow of remittances reverts back to its projected normal level. Inflow of workers remittance has been playing a very important role in the growth of national savings in Bangladesh in recent years. The increase in national savings projected under the Plan will depend on the continued growth in remittances. However, given that agriculture remains an important dominant economic sector of the country, savings behaviour of farming sector of the country will be an important determinant of overall savings of the country. This report investigates the saving behaviour of the farming sector of the country. In the next section the report describes the agriculture sector of the country, followed by analysis of household savings, based on survey of household income and expenditure. 2. THE ECONOMY, AGRICULTURE AND FOOD SECUIRITY IN BANGLADESH 2.1 Macro Economic Scenario The GDP of Bangladesh over the last five years grew consistently at an average rate above 6 percent. The growth of import-export trade, secular increasing trend in the flow of remittance and the favourable balance of the current account kept the external sector to some extent stable. 6

Fig. 2 Recent changes in structure of GDP During the recent decade overall GDP of Bangladesh has been considerably shown upward trend. But the growth in agriculture GDP slightly declined with an average of 3.4% during 1997 to 2008. Agriculture being the supportive of growth of the economy, there is no other alternative but to develop agriculture sector for alleviation of poverty by attaining accelerated economic growth. Since provision of food security, improvement of the living standard and generation of employment opportunities of the huge population of the country are directly linked to the development of agriculture, there have been continued efforts by the Government for the overall development of this sector. 2.2 Sectoral overview of agriculture Agriculture is the most important sector of Bangladesh economy due to its role in food security, employment and livelihood. The current share of agriculture to GDP is around 21%, although this share has been declining in the last ten years along with the one of the agricultural sub sectors (Table 1). Still more than 70% of the people in Bangladesh are directly or indirectly employed in this sector. The agriculture of Bangladesh is dominated by crops which accounts about half of total agricultural GDP Fisheries & livestock are also important sub-sector contributing to agricultural GDP (Table 1). It is important to note that in the recent decade the subsector of livestock, forestry and fisheries had faster growth rate (Table 2). There are, however, substantial year to year fluctuations in these rates and it is most pronounced in case of crops. Table 1.Contribution of agriculture to GDP (%) at constant prices (Base: 1995-96=100) Sector/ Sub- Sector 1999-00 2000-01 2001-02 2002-03 Year 2003-04 2004-05 2005-06 2006-07 2007-08 Agriculture 25.58 25.03 23.99 23.47 23.08 22.28 21.85 21.37 20.88 A. Crops 14.59 14.7 13.75 13.43 13.23 12.51 12.28 12 11.7 B. Livestock 3.02 2.95 2.96 2.93 2.91 2.95 2.92 2.88 2.79 C. Forestry 1.88 1.87 1.88 1.86 1.83 1.82 1.79 1.76 1.75 D. Fishing 6.09 5.51 5.4 5.25 5.11 5 4.86 4.73 4.64 Source: Statistical Yearbook of Bangladesh 2007 7

Table 2.Growth rate of agriculture and its sub-sectors at constant price (Base: 1995-96=100) Sector/ Sub- Sector 1999-00 2000-01 2001-02 2002-03 Year 2003-04 2004-05 2005-06 2006-07 2007-08 GDP growth rate 5.94 5.27 4.42 5.26 6.27 5.96 6.63 6.43 6.21 Agriculture 6.16 4.59 2.36 3.54 4.13 4.03 5.07 4.81 3.85 A. Crops 8.10 6.18-2.39 2.88 4.27 0.15 5.03 4.43 3.44 B. Livestock 2.74 2.81 4.70 4.51 4.98 7.23 6.15 5.49 2.41 C. Forestry 4.94 4.85 4.91 4.43 4.18 5.09 5.18 5.24 5.42 D. Fishing 8.87 4.53 2.22 2.33 3.09 3.65 3.91 4.07 4.11 Source: Statistical Yearbook of Bangladesh 2007 & Statistical Pocket Book 2008 Rice dominates Bangladesh agriculture covering more than 80% of the land area. The production of main staple rice has shown a long term growth trend of 2.8 percent per annum over the period from 1981/82 to 2006/07. During 1997 to 2005, total rice acreage changed little, T. Aman acreage remained almost unchanged, while irrigated Boro acreage substantially increased with the reduction of rain-fed Aus which showed about 6.3 percent annual growth during the same period. Boro rice accounts for about 60 percent of total food grain production (Figure 3 and 4). Figure 3. Rice acreage by season Figure 4. Rice output by season During the same period production of potato and vegetables also increased considerably (Fig 5). For other crops long term growth remains unsatisfactory. Usually little land is used to grow nutritionally important food crops such as vegetables, fruits, pulses and oilseeds (Table 3). During recent years both production and area of wheat, pulses and oilseeds have been greatly reduced. 8

Production in th 35000 30000 25000 20000 15000 10000 5000 0 Rice Wheat Potato Pulses Oilseeds Vegetables Fig. 5 Trend in production of rice and non-rice crops Table 3 Percentage of land area planted to different crops in 2004-05 Land area percentage Rice 79.4 Wheat 4.09 Pulses >2.80 Oilseeds >2.37 Potato >2.39 Jute >2.86 Fruits >1.36 Sugar cane >1.15 Tea >0.39 Spices and condiments >2.21 Vegetable >2.14 During recent years both production and area of wheat, oilseed and pulses have greatly reduced. The growth of pulses was only 0.3% while sugar was negative 1.2% other than rice and potatoes all other crops showed the decline in growth rate (Table 4). Table 4 Growth of non rice crops in Bangladesh during 1971 to 2007 Food Items Percent/Year (1971-2007) Wheat -2.9 Pulses 0.3 9

Oil crops 2.3 Potatoes 3.0 Vegetables 1.5 Sugar -1.2 Fish 2.5 Meat 2.3 Milk 2.3 Egg 4.0 The preceding section describes the contribution of principal drivers in crop sector growth: Irrigation The recent agricultural growth in the country has been largely due to expansion of minor irrigation through the use of Deep Tubewells (DTWs), Shallow Tubewells (STWs) and Low Lift Pumps (LLPs) (Figure 6). During 2006, there were 29,170 DTWs, 12,02,720 STWs and 1,07,290 LLPs fielded and the total irrigated area was estimated at 4.883 M ha which was 56.51% of the net cultivable area of the country; where irrigation coverage by groundwater and surface were 80.6% and 19.4%, respectively. Again, in 2007 the national irrigation coverage was 5.05 million hectares which is 60.92% of the total cultivable land, where groundwater covered 78.98% and surface water covered 21.02% of the total irrigated area (MoA, 2008). Fertilizer use Fertilizer use in Bangladesh has increased from 3.28 million tons in 2001-02 to 3.55 million tons in 2006-07 (Figure 4). During 2002 to 2009 the growth rate of total fertilizer use in Bangladesh was 4.2% per annum. Between 2002 and 2009, the demand for fertilizer in Bangladesh has increased by 25%. Fertilizer prices in Bangladesh have risen very sharply during 2008. The cost of Triple Super Phosphate (TSP) and Muriate of Potash (MOP) rose from BDT 16 in 2007 to BDT 50 and BDT 40 per kg, respectively while the cost of urea was maintained at BDT 6 per kg by a government subsidy, Area Irrigated in 000' ha 5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 : 82/83 83/84 84/85 85/86 86/87 87/88 88/89 98/90 90/91 91/92 92/93 93/94 94/95 95/96 96/97 97/98 98/99 99/00 00/01 01/02 02/03 03/04 04/05 05/06 06/07 07/08 Year Fig.6 1: Yearly Trend Trend of Irrigation in Irrigated Growth Land in Bangladesh Area in Bangladesh 10

reduced greatly by government. It is however noted that very recently the government has increased price urea to Tk 12 per kg and the prices of MOP and TSP have been largely reduced through government intervention. In the fiscal year 2010-11, government has decided to support 1.82 crore farmer families under the agro-input Assistance-Card for purchasing diesel, fertilizers, seeds and receiving credit. Supply of Quality Seeds A number of agencies, both public and private are presently involved in seed production and distribution system of the country. The fact is that only about 18% seed requirements of the country can be met from certified and Truthfully Labelled Seeds (TLS) of Government and private sources, the remaining 82% seed comes from the farmers own seed. There are serious problems of seed quality supplied by public, private and farmer s saved seed. Livestock In 2006-07, total cattle population of the country was about 23 million, buffaloes 1.21 million, goat and sheep 23 million and poultry and duck 246 million. The production of milk, meat (beef, mutton and chicken) and eggs had an increasing trend over the past several years (Fig 7 and 8). However, its total production is still far below the national requirements. The milk production showed slow growth but sustained growth while recently meat production is slightly decreasing. On the other hand, the egg production largely fluctuates due to avian influenza. 30 25 Production (Lac ton) 20 15 10 5 0 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 Milk Meat Egg production (Lac Number) 58000 56000 54000 52000 50000 48000 46000 44000 42000 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 Fig. 7 Trend in production of milk and meat in Bangladesh Bangladesh. Data from Bangladesh Economic Review, 2009 Fig. 8 Trend in egg production in Fisheries Long term growth rate in production of fish during 1971 to 2007 was 2.5%. During 2001-02 to 2007-08 total fish production increased from 1.89 million MT to 2.59 million MT with an average annual growth rate of 5.3%, while inland fish and cultured fish production had growth rates of 8.2% and 4.2%, respectively (Table 5). Overall shrimp production has increased steadily over the last 20 years (Figure 9 and 10), but still much lower than that of the neighbouring countries such as Thailand with 800 kg/ha (Samsak et al. 2005) and India with 600 kg/ha (Vasudevappa, 2005). Among shrimp producing 11

countries, Bangladesh ranks fourth with respect to area under shrimp farming and sixth in volume of production. 30 25 Production in Lac metric ton 20 15 10 5 0 2001-02 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 Inland fish Culture fish Marine fish Total fish Fig 9. Trend in fish production Fig. 10. Trend in shrimp production Data from Bangladesh Economic Review, 2009 Table 5. Recent trend in fish production in Bangladesh during 2001-08 Commodity Growth (during 2001-2008) Inland fish 8.2 Culture fish 4.2 Marine fish 2.4 Total fish 5.3 Source: Bangladesh Economic Review, 2009 and authors calculation Forestry The production of wood in Bangladesh since 1970 is increasing. This increase depends largely on the expansion of new social forestry plantations involving private participants. Beside these, over 10 million homesteads in about 88 thousand villages across the country possess a large quantity of trees. These homesteads are the major suppliers of wood for the nation. The role of the homesteads in growing stock is presented in Table 6. The total forest area in Bangladesh is 2.53 million hectare in 2007; but all do not necessarily carry tree cover. The tree covers 48.8% of the forest land. Table 6. Estimated growing stock of wood in Bangladesh Forest type Gross volume of growing stock of wood, 12

million m 3 1990 2000 2005 Hill forest* 19.34 17.26 16.03 Plain land forest* 2.07 1.98 1.97 Littoral forest* (Mangroves) Village/homestead forest** 13.14 11.97 12.00 69.16 89.58 101.94 * Data from FAO, 2005. **Hammermaster data of 54.8 million m 3 in 1981 with growth rate of 2.62 ( Choudhury and Hossain 2009 used to derive the estimated data). 2.3 Current Food security situation About 40% of Bangladesh s total population are calorie poor, consuming less than 2122 kcal per capita per day, while one-fifth of population are hard core poor who consume less than 1805 kcal per capita per day (BBS, 2007). According to SOFI 2009, Bangladesh has an estimated 40.2 million undernourished people in 2004-06 1. One positive aspect is that the proportion of undernourished in total population has declined from 36% in 1990-92 to 26% in 2004-06. Out of a total of 336.6 people who are estimated to be undernourished in South Asia, 12% comes from Bangladesh (India s share is 75%). The most recent estimates (Household Income Expenditure Survey HIES, 2005) revised to up to 56 million (still 40% of the population) the absolute number of undernourished people, those who fail to meet minimum level of caloric consumption needs of 2122 Kcals/person/day. Within this population 27 million are unable to consume 1805 Kcal/person/day and 11 million are below the consumption level of 1600 Kcal/person/day. Food security worsens with inter-year shortfall in food grain production caused by climatic variations and natural disaster such as floods, tidal surge and insect and pest attacks. Variations in food intake also exist between regions of the country, between adults and children and between men and women at the household level. Household Income and Expenditure Survey 2000 shows that malnutrition is widespread among the poorest 14 percent of the rural population, who consume only about 1600 Kcal per capita per day. Another 10 percent consume between 1600 and 1800 Kcal (BBS, 2000). Bangladesh Demographic and Health Survey (2004) shows that 43 percent of children under 5 are stunted and the prevalence of stunting increases from 10 percent of children below 6 months to 51 percent of children from 48-59 months of age. The survey also shows that mothers suffer from calorie deficiency and pregnant women suffer from anaemia. 1 Undernourishment exists when caloric intake is below the minimum dietary energy requirement (MDER). The MDER is the amount of energy needed for light activity and a minimum acceptable weight for attained height, and it varies by country and from year to year depending on the gender and age structure of the population. 13

Bangladesh has made good progress since 1992 in reducing income poverty based on the national poverty line. The country was able to lower the overall incidence of poverty from 58.8 percent in 1991-92 to about 48.9 percent in 2000, with an annual rate of decrease of 1.8% per year. It further declined to 40% in 2005 with a decreasing rate of 3.9% per annum (Fig. 11). In spite of the advancement, about 57.7 million people are poor with one-third caught in hard-core or extreme poverty. Fig 11. Proportion of population below national poverty line during 1991 to 2005 (estimates for 2007) Bangladesh past efforts towards improving food security have resulted in a 5% point decrease in the proportion of undernourished between 1990-1992 and 2002-2004 down to 30%. However, mainly due to population growth, the number of undernourished actually increased from 39.3 million to 44 million (POA, 2008-15). Based on one of the population projection of 181.4 million in 2015, attaining the MDG hunger target implies that the number of undernourished people must decrease to 31.8 million. Meeting the WFS hunger, i.e. reducing the number of undernourished to 19.65 million by 2015, implies that the MDG1 target needs to be overachieved with the proportion of the undernourished brought down to 11% (POA, 2008-15): a significant challenge for Bangladesh where the absolute number of undernourished has seen only a limited decrease in the period 2005-2009. Table 7: Proportion and number of undernourished in Bangladesh Past status Current Mid/Long Term Targets 1995-97 2003-05 2004-06 2015 FAO Proportion of undernourished (%) Baseline 40 27 26 18 (MDG) 14

1990-92: 36 FAO- Number of undernourished, million Baseline: 1990-992: 41.6 51.4 40.1 40.2 20 (WFS) Source: Monitoring Report 2009 of the National Food Policy Plan of Action. Total food grain (rice and wheat) production in 2006/07 stood at 28.05 million tons, but there was a short fall of about 1.05 million tons of rice (Aus and Aman) in 2007/08 due to double floods in August and September 2007. The total food grain production further increased to 32.17 million in 2008-09 (Monitoring Report, 2009 Fig 12, below). In 2007-2009, rice availability, defined as the sum of net production and total import, was estimated to be 14-15% higher than the estimated human consumption needs- assuming a consumption of 453 gm/person/day- while net production alone was estimated to be 12.6% in exceeds of needs implying achievement foodgrain self-sufficiency. Availability of pulses, oilseeds, vegetables and fruits, which are the chief sources of protein, mineral and vitamin still remain far below the actual requirements, making it difficult to provide adequate diet for all. Fish production increased from 0.95 million tons in 1991-92 to 2.33 million tons in 2005-06. Meat, milk and egg production has also increased significantly over the years (Table 5), but the availability of these food items is far less than required for a nutritionally balanced diet. Fig 12. Rice production in Bangladesh (from 1972-73 to 2008-2009) Generally, improvements in the production/availability of food grains have not been accompanied by commensurate increases in the production of other major foods, such as, pulses, oilseeds fruits, vegetables, fish, meat and eggs, which are essential for a diversified and balanced diet. For a large growing population of Bangladesh, it is more important to know the per capita availability of food items. On a per capita basis, availability of cereals has increased from 453 gm/day in 1991/92 to 540 gm/day in 2006/07. Per capita availability of potato, fruits and vegetables has also increased, but per capita availability of pulses and oilseeds has remained stagnant or declined (Table 8). Fish availability per capita increased from 27 gm in 1994-95 to 41 gm in 2003-04. Availability of meat, milk and eggs per capita has also increased. Per capita per day availability of energy has increased from 2069 calories in 1991/92 to 2489 calories in 2003/04. Table 8: Per Capita Availability of Major Food Items: 1994-95 to 2004-05 15

Availability Food Items (gm/capita/day) 1994-95 1999-00 2004-05 Potato Pulses 32 11 57 8 108 10 Oilseed 10 8 10 Vegetable 21 24 108 Fruits 24 22 68 Fish 27 35 41* Meat 11 15 21 Milk (ml) 35 36 45 Egg (million) 2.76 4.20 Source: BBS, DAE, DLS, DOF, BARC; * Refers to 2003-04. 5.54 The average Bangladeshi diet contains very high intake of cereals (74.4% of total dietary energy of 2240 kcal) with too little contents of non-cereal food items and thus an average diet is not only deficient in energy but also imbalanced and deficient in vitamins and minerals. Gaps between requirement and production of other important food crops and livestock products are wide. According to MOA estimate, deficit of pulses and oilseeds are 1.12 million tons and 2.26 million tons respectively (2004-05). Similarly, shortage of vegetables and fruits is estimated at 2.24 million tons and 2.70 million tons respectively. The national deficit of milk and meat is 10.54 million tons and 5.12 million tons respectively, and of eggs is 9,253 million (DLS, 2007). The national requirement of fish is estimated at 3.0 million tons, showing a deficit of 0.67 million tons (DOF, 2007). There is wide gap between projected cereal production and requirements for 2014-15. The implications of this projected consumption and requirements for acceleration and sustenance of cereal and non-cereal food production is that it is necessary to change in land use pattern for crop diversification, and improvement in food processing and food consumption habit. 3. Investment in agriculture 3.1 Overall investment in agriculture There is no available time series data of investment in the agriculture sector of Bangladesh. However, BBS calculated GFCF in agriculture for the 4 years only from FY 97 to FY 00. Table --- gives total investment in the economy and in agriculture, as estimated by BBS 2. Table 5: Investment in economy and in agriculture (Million Taka) FY 97 FY 98 FY 99 FY 00 Total 242,427 275,863 315,513 355,149 Public 91,164 102,851 118,279 134,247 Private 151,263 173,012 197,234 220,902 Investment in agriculture (total) 20,025 22,924 26,409 29,904 Public 4,657 4,793 5,785 6,825 2 Methodology used to calculate GFCF is not publicly available nor any explanation why this attempt was not continued. 16

Private 15,368 18,131 20,624 23,079 Source: BBS From the table, we can note that the average growth rate of public Investment is around 13.7 percent while average growth rate of private investment is around 13.4 percent. Investment in agriculture increased at a rate of around 12.5 percent per year. Both public and private investment in agriculture increased at about the same rate of the total. However, more than 75 percent of total agricultural investment is from private sector and this share remained same during the estimated period. Shamsur Rahman and Atiqur Rahman (2002) also estimated the level of investment (capital formation) in Bangladesh agriculture. According to their estimates, while the investment in the Service sector grew exponentially, investment in agriculture remained static. Figure ---: Trends in major sectoral capital stock in Bangladesh, 1980/81 to 1998/99 (million taka) 4500000 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 Service Industry Agriculture Data Source: Shamsur Rahman and Atiqur Rahma (2002) 3.2 Public Investment As regard public investment, the allocation of ADP to Ministry of Agriculture (MOA) is a good indicator of the public investment in and for agriculture. Table --- gives ADP allocation of Ministry of Agriculture from 1980-81 to 2006-07. It shows that though the share of allocation to Ministry of Agriculture has declined from about 16 percent in 1980 to about 2 percent in 2007, the absolute figures have increased moderately. In 1980, the allocation was 3729 million taka and in 2007 it was 5507 million taka. MOA invested almost entire amount on crop and irrigation. Table 11: ADP allocation of Ministry of Agriculture from 1980-81 to 2006-07 (Million Taka) Share Country of Plan Period Year Crop Minor Irrigation Forestr y Rural Institution Total Allocation MOA (%) Second five year 1980-81 1834.3 0 1895.60 - - 3729.9 0 23689.90 15.74 1981-82 2096.4 0 1925.60 - - 4022.0 0 27152.50 14.81 1982-83 2504.7 0 1668.10 - - 4172.8 0 31263.00 13.35 1983-84 2894.7 0 1831.60 - - 4726.3 0 35847.00 13.18 1984-85 2096.1 0 1661.00 - - 3757.1 0 35084.10 10.71 17

Plan Period Year Crop Minor Irrigation Forestr y Rural Institution 1985-1154.7 Third five year 86 0 1195.60 - - 1986-1351.6 87 0 1310.20 - - 1987-1425.5 88 0 1137.70 - - 1988-2001.5 89 0 1431.00 - - 1989-1466.0 90 0 1427.40 - - 1990-1898.0 Fourth Five Year 91 0 1192.00 - - 1991-2892.3 92 0 1332.00 - - 1992-2276.0 93 0 1096.20 - - 1993-2780.4 94 0 1302.00 - - 1994-2942.9 95 0 1490.40 - - 1995-2952.6 Two Year 96 8 1204.30 - - 1996-3434.6 97 5 952.00 - - 1997-3589.3 Fifth Five Year 98 0 721.80 - - 1998-3709.7 99 0 810.28 - - 1999-4292.9 00 0 903.40 - - 2000-3713.1 01 0 974.50 - - 2001-3346.6 02 0 847.10 - - 2002-3409.7 Three Year 03 8 817.70 - - 2003-3677.3 Rolling Plan 04 0 1030.30 23.40 200.00 2004-3123.2 05 0 1542.10 200.00 2005-3363.3 06 0 2430.70 200.00 2006-3354.6 07 0 1923.20 29.20 200.00 Source: Ministry of Agriculture Share Country of Total Allocation MOA (%) 2350.3 0 40960.00 5.74 2661.8 0 45130.00 5.90 2563.2 0 46510.00 5.51 3432.5 0 45960.00 7.47 2893.4 0 51030.00 5.67 3090.0 0 61210.00 5.05 4224.3 0 71500.00 5.91 3372.2 0 81210.00 4.15 4082.4 0 96000.00 4.25 4433.3 113500.0 0 0 3.91 4156.9 104470.0 8 0 3.98 4386.6 117000.0 5 0 3.75 4311.1 122000.0 0 0 3.53 4519.9 140000.0 8 0 3.23 5196.3 165000.0 0 0 3.15 4687.6 182000.0 0 0 2.58 4193.7 160000.0 0 0 2.62 4227.4 171000.0 8 0 2.47 4931.0 190000.0 0 0 2.60 4865.3 220000.0 0 0 2.21 5994.0 245000.0 0 0 2.45 5507.0 260000.0 0 0 2.12 4. Savings behaviour of the farming households 18

4.1 Ownership of Land Table 13 presents distribution of households by size of land owned and operated in rural area of Bangladesh. It is revealed from the table that, in rural area, percentage of household having no land reduced to some extent in 2010. Household having land up to 0.49 acre increased from 55.2% in 2005 to 60.5% in 2010, whereas, it was 60.0% in 2000. It is also reveled from the table that, the percentage of households with land ownership 7.50 acre and over decreased to 1.1% in 2010 compared to 1.6 in 2005 and 1.3 in 2000. In case of operated land, it may be seen from the same table that, the percentage of households having smaller size of operated land i.e. up to 0.49 acre increased over time. It was 52.9% in 2005 which increased to 55.4% in 2010. Table 13: Rural Household by Size of Land (Owned and Operated) Land Size (in acre) HIES-2010 HIES-2005 HES-2000 1 2 3 4 Owned Land Total 100.0 100.0 100.0 Landless 4.6 5.3 5.6 0.01-0.49 60.5 55.2 60.0 0.50-0.99 11.6 12.4 9.5 1.00-2.49 14.6 17.6 15.4 2.50-7.49 7.6 8.7 8.1 7.50+ 1.1 1.6 1.3 Operated Land Total 100.0 100.0 100.0 Landless 3.6 4.1 4.9 0.01-0.49 55.4 52.9 54.0 0.50-0.99 14.2 14.2 11.9 1.00-2.49 18.3 19.1 18.8 2.50-7.49 7.8 8.6 8.8 7.50+ 0.7 0.9 1.6 4.2 Distribution of household by occupation Table 14 provides distribution of head of households by main occupation and by residence. In 20I 0, the highest 36.1% of the head of household was engaged in agriculture, animal husbandry, forestry and fisheries, I6.9% were engaged in production & related works and transport works, 11.6% in the sales occupation and 2.9% were in clerical and related workers and government executives. The percentage of head of household engaged in administrative and managerial works was 2.9%. In the rural area, 46.4% of the head of households engaged in agriculture occupation followed by production and related workers and transport works 13.4%. In the urban area, 26.5% households are engaged in production and related works and transport workers followed by 18.9% in sales works. Table 14: Distribution of Head of Households by Residence and Main Occupaiton-2010 Sly No. Major Occupation Nationa Rural Urban I 1 2 Professional, technical & related workers 3 6.1 4 4.8 5 9.5 2. Administrative & managerial works 2.9 1.9 5.8 3 Clerical & related works and govt. executive 2.9 1.3 7.0 4 Sales workers 11.6 8.9 18.9 5 Service workers 5.9 5.4 7.4 7 Agricultural, animal husbandry, Forestry & 36.I 46.4 8.3 19

8 Production & related workers and transport 16.9 13.4 26.5 9 Head not working I7.6 I7.9 16.7 10 Total 100.0 100.0 100.0 4.3 Level of Income and Expenditure by Size of Own Land in Rural Area Table 15 provides information on monthly household income, expenditure, family size and number of earners by size of own land in the rural area. In the landless group, the average income per household was taka 5713 in 2010. The corresponding figures for 2005 & 2000 were taka 3299 & 3248 respectively. On the other hand, the average income of the households owning land size 7.50 acres & above was taka 29673 which was 5.2 times higher than the average income of landless group. Average monthly income per household in rural area increased with the increase of size of owned land owned. Thus, land holding size is an important determinant of income particularly in rural area. It appears from the table 15 that, the highest percentage of households (38%) owned land 0.05-0.49 acres, where 80% of households owned land 0.01-1.49 acres. This indicates that, the farm size is very small in Bangladesh. Another important feature was found that, the family size increased with the increase of own land size owned. The family size were 3.83 for the landless, 4.09 for land size 0.01-0.04 acre, 4.53 for land size 0.05-0.49 acre, 4.69 for land size 0.50-1.49 acres, 4.88 for land size 1.50-2.49 acres, 5.37 for land size 2.50-7.49 acres and 5.83 for land size 7.50 acres or more. The number of earners per household also increased with the increase of land size except for landless household. Number of earner was 1.27 in rural area where 1.31 for the landless, 1.16 for land size 0.01-0.04 acre, 1.26 for land size 0.05-0.49 acre, 1.27 for land size 0.80-1.49 acres, 1.34 for land size 1.50-2.49 acres, 1.50 for land size 2.50-7.49 acres and 1.83 for land size 7.50 acres or more. The monthly household expenditure by own land size also shows increasing pattern with the increase of Own land size owned. Table: 15 Income, expenditure and savings by holding sizes Holding Size in Acres % of Holdings Income Expenditure Savings % of income saved All Group 100 9648 9612 36 0 Landless 5 5713 6507-794 -14 0.01-0.04 23 5973 6735-762 -13 0.05-0.49 38 8602 9010-408 -5 0.50-1.49 19 10785 10518 267 2 1.50-2.49 7 13198 12424 774 6 2.50-7.49 7 19147 16035 3112 16 20

7.5+ 1 29673 24457 5216 18 If income-expenditure, i.e. non-consumption, is counted as saving then, overall, based on a nationwide survey, it appears that in Bangladesh the farming household in aggregate saves very little. Household income and expenditure survey shows that, as indicated in Table 15, for all group of households, the percentage of income saved is almost nil. Household with no land and those with land but with holding size of less than half a acre, nearly 66 percent of the holdings of the country, have negative savings. The holding size of up to 1.5 acre also do not have any positive savings. Therefore if this group is also included then it would appear that all households with less than 1.5 acre of land, amounting to nearly 85 percent of the farming households, including the landless, do not have any positive savings. Only the households with holdings of 1.5 acre and above generate positive savings, and higher the holding size higher the rate of savings. However, the situation is different once we consider some types of expenditure, such as those on housings and purchase of durables as investment. 5. Saving behaviour of households with and without migrant workers 5.1 Surveys of Savings and investment by Household with and without migrants A field survey was conducted to assess household level savings and investment. The size of the sample for the survey was 418 and it covered all of the 7 Divisions of Bangladesh. The districts selected for the survey were Mymensingh, Comilla, Pabna, Jessore, Patuakhali, Maulavibazar and Thakurgaon. From each of the district one random union was selected for the survey. From each union sixty households were selected for the survey (except one union, where the number was 58). About 85 percent of the households in the sample have less than 7 members (Table 16). About 70 percent (Table 17) members of the households are literate and only 33 percent among them have secondary education and above. Table 16: Average Family Size Family Frequenc Percentage size y 1-2 10 2.47 3-4 139 33.30 5-6 205 49.00 7 and above 64 15.30 Table 17: education level 21

Level of Frequency Percentage Education Illiterate 123 29.4 Primary 114 27.3 Secondary 138 33 Higher secondary and above 42 0.20 What emerged is that in terms of savings and investment that there is clear distinction between the families with a migrant and those without a migrant. Average annual household saving for migrants amounts to Tk.87583 compared to Tk. 28957 for non-migrant households. Thus migrant households save about 25 percent of their total annual income while non-migrant households save about 22 percent of their total annual income. There is large difference between absolute saving of these two groups of households; the difference between the proportions of income saved is small. Savings are held in different forms, the most popular form being bank deposits. This was also confirmed by participants in FGDs where they revealed that they maintain bank accounts for depositing remittance. Whenever some fund is needed taka is withdrawn from the account leaving the rest in the account. Earlier studies estimated migrant households savings at 28 45 percent of remittance. Our estimate is somewhat lower than the earlier studies for two reasons. First, the saving rate is for agricultural households and not for all types of households. Second, saving rate is calculated as proportion of total income instead of remittance only. Since expenditure is fungible, it is more meaningful to estimate saving as proportion of total income. Agricultural households both migrant and non-migrant, invest in agriculture as well as outside agriculture. Over the last years migrant households invested Tk.322381 compared to Tk. 225179 by non-migrant households. The highest amount of income has been invested in house construction by migrant households; the same holds for non-migrant households. This is a general feature of household behaviour live in better house as income increases. An interesting finding is that of higher investment in agricultural equipment by migrant households and the difference is statistically significant Table 18 Average annual household savings in Taka Migrant Non Migrant Mean Difference t Sig. Bank deposits 89530.7 3 31044.0 3 58486.70 3.770 0.000* Post office 31333.3 20000 11333.33 1.205 0.351 22

NGO Others Total 3 24430.4 3 70785.8 7 87582.7 4 15892.7 3 8537.708 1.004 0.324 20753.8 5 50032.02 1.019 0.316 28957.0 4 58625.70 4.267 0.000*. Table 19 Average household investment portfolio in Taka over last 10 years Investment heads Migrant Non Migrant Mean Difference Land Improvement 67750 36819.72 30930.28 House Construction 471500 327430.6 144069.4 Business start/ expansion 208547.6 184750 23797.62 Purchase of draught animal 64588.89 44947.37 19641.52 Purchase of breeding animal 30853.7 29472.97 1380.731 Purchase of agri equipment 22280 6361.111 15918.89 Purchase of Transport 133660 130978.6 2681.429 Total 322381.2 225179.3 97201.87 Agricultural households hold real as well as financial assets. They also reduce their financial obligation. As expected the migrant households spend a large amount on land. It was pointed out during FGD that migrant households are not limited to purchase of land in their own village or locality; they also buy land in urban areas. Acquisition of assets by migrant households is significantly different (higher) from that of non-migrant households in most cases. Table 20 Average household assets in Taka over last 10 years Migrant Non Migrant Mean Difference Purchase of Land Improvement of House Payment of Loan 558364. 3 205550 352814.3 232105. 3 43750 188355.3 247501. 9 190916.7 56585.19 23

Purchase of transport for own use Purchase of gold, silver 58970.8 3 17150 41820.83 94576.6 7 50540.54 44036.13 Others 25000 23333.33 1666.667 Total 432715. 6 185300 247415.6 5.2 Focus group discussion Focus group discussion was arranged in three villages in three districts to assess the impact of migration on poverty and food security and agriculture and rural development. These are Chapil of Dhaka district, Rashid Deohata of Tangail district and Dhonuakhola of Comilla district. In Chapil about 60 percent of the households have migrant workers, in Rashid Deohata about 18 percent and in Dhonuakhola about 75 percent of households have migrant workers. An average of 10 persons including women from migrant agricultural households and 5 persons from non-migrant agricultural households participated in the discussions. Besides, owners of tractors and power pumps in the villages were separately met. Migrant workers come from households which can bear the cost of migration either from own savings or borrowing or both. Thus households without any savings or assets such as land to sell or mortgage can hardly send workers abroad. However, it is reported that sometimes migrant workers facilitate migration of their relatives or neighbours which makes migration even from less well off households possible. This is how 75 percent of households in Dhonuakhola in Comilla have been able to send their members abroad. 6. Household income and expenditure based on household income and expenditure survey 2010 This chapter details with estimates of household income and expenditure based on the findings from Household Income and Expenditure Survey 2010. Income and expenditure estimates are central to the entire findings of the Survey. Moreover, distribution of income and expenditure by decile group, Gini co-efficient by income, expenditure by land ownership, consumption expenditure and expenditure by major food items have also been discussed in this chapter. The economic scenario presented here is the fundamental basis for short and long term economic analysis of the economy. 6.1 Level of Income Table 21 provides monthly income per household, number of earner per household, monthly income per member (monthly per capita income) and monthly income per earner etc. Average monthly income per household at current price was estimated at taka 11479 at the national level in 2010. This was Tk. 7203, Tk. 5842 and Tk. 4366 in 2005, 2000 and 1995-96 respectively. In 2010, the monthly household income increased by 59.36% when compared with 2005 and by 162.9% with respect to 1995-96. Per capita monthly income was estimated at taka 2553 in 2010. This was Tk. 1485, Tk. 1128 and Tk. 830 in 2005, 2000 and 1995-96 respectively. 24

Per capita monthly income increased by Tk. 1068 (71.92%) in 2010 compared to 2005 and increased by Tk. 1723 (207.59%) over the year 1995-96. It is notable that in 2010, rural income increased at a higher rate than that of urban income. In 2010, rural income increased by 58.29% whereas urban income increased by 57.45% in 2010. This increase of income in both urban and rural areas accelerated the process of poverty reduction in both urban and rural areas. 25

Table 21: Number of Members, Earners, Household Income, Income Per Member and Earner By Residence Survey year & residence Number of members per household Number of earners per household Monthly household income per household Monthly income per member Monthly income per earner 01 02 03 04 05 06 National 2010 4.50 1.31 11479 2553 8795 2005 4.85 1.40 7203 1485 5145 2000 5.18 1.45 5842 1128 4029 1995-96 5.26 1.48 4366 830 2950 Rural 2010 4.53 1.27 9648 2130 7592 2005 4.89 1.37 6095 1246 4449 2000 5.19 1.43 4816 928 3368 1995-96 5.25 1.46 3658 697 2505 Urban 2010 4.41 1.40 16475 3740 11778 2005 4.72 1.50 10463 2217 6975 2000 5.13 1.54 9878 1926 6414 1995-96 5.30 1.59 7973 1504 5014 Urban as % of rural 2010 97 110 171 176 155 2005 96 109 172 178 157 2000 99 108 205 208 190 1995-96 101 109 218 216 200 In HIES-20 I 0, the number of earners per household was found to be 1.31 at national level, 1.27 in rural area and 1.40 in urban area. Since, 1995-96, number of earner per household shows a declining trend and has a positive correlation with the declining changing pattern of household size. 26

In 2010, income per earner was found to be taka 8795 for the country as a whole. In rural area, this was taka 7592 and in the urban Tk. 11778. Income per earner increased to Tk. 8795 in 2010 from Tk. 5145 in 2005, an increase of Tk. 3650 (70.94%) during this period. 6.3 Income Distribution Decile groups from the table 22 shows the pattern of distribution of income from percentage share of income of households. It is evident from the table 4.2 that the gap between the poorest of the poor (bottom 5%) and the richest of the rich (top 5%) is extremely high. In HIES 201 0, the income accruing to top 5% of the household was 24.61%, whereas, the same was 0.78% for the bottom 5% household. In 2005, income accruing to top 5 percent of the households was 26.93%, where the same was 0.77% for the bottom 5%. It slightly increased from 0.77% to 0.78% in 2010. On the contrary, the share of income of the top 5% decreased from 26.93% to 24.61% over the same period indicating redistribution of income in favour of the mid-level. It is also evident from the table 22 that, income accruing to households belonging to Decile-] to Decile-5 remains almost same in 20 10 and recorded at 2.00%, 3.22%, 4.10%, 5.00% and 6.01% respectively at national level. Percentage shares in 2005 were 2.00%, 3.26%, 4.10%, 5.00% and 5.96% respectively. These five deciles jointly share only 20.33% of total income, although they comprise 50% of the population. In 2010, the income share of the households belonging to Decile-6 to Decile-9 increased but, that of Decile-10 decreased corresponding to those of2005. In 2010, the decile-6 shared 7.32% of the total income, decile-7 9.06%, decile-8 11.50%, decile-9 15.94% and decile- I 0 35.85%, whereas, these were 7.17%, 8.73%, 11.06%, 15.07% and 37.64% respectively in the year 2005. For both rural as well as urban area, show the similar changing pattern of decile distribution of income as the national level. Table 22 Percentage share of income of households by Decile group and Gini Co-efficient Household Income Group 2010 2005 Total Rural Urban Total Rural Urban 01 02 03 04 05 06 07 National 100.00 100.00 100.00 100.00 100.00 100.00 Lower 5% 0.78 0.88 0.76 0.77 0.88 0.67 Decile-1 2.00 2.23 1.98 2.00 2.25 1.80 Decile-2 3.22 3.53 3.09 3.26 3.63 3.02 Decile-3 4.10 4.49 3.95 4.10 4.54 3.87 Decile-4 5.00 5.43 5.01 5.00 5.42 4.61 Decile-5 6.01 6.43 6.31 5.96 6.43 5.66 Decile-6 7.32 7.65 7.64 7.17 7.63 6.78 27