Journal of Agricultural Economics and Rural Development

JAERD

Vol. 2(2), pp. 032-038, July, 2015. www.premierpublishers.org, ISSN: 2167-0477

Research Article

Diversification of rural livelihoods in Bangladesh

1*

Md. Tanvir Ahmed, 2Humnath Bhandari, 3Prudenciano U. Gordoncillo, 4Cesar B. Quicoy and
Gideon P. Carnaje

1*,3,4,5

Department of Agricultural Economics, University of the Philippines, Los Banos, Philippines.

Scientist, Social Sciences Division, International Rice Research Institute, Philippines.

Diversification of livelihood activities minimizes households vulnerability to shocks by

reducing income variability. Livelihoods of rural Bangladesh is encompass both farm and nonfarm activities. This study investigated the patterns and extent of diversification of livelihoods in
rural areas of Bangladesh. The study drew a random sample of 500 rural farm households in
Bangladesh through a multi-stage random sampling technique. The primary data were collected
using structured questionnaires coupled with interview schedules and the data were analyzed
using descriptive statistics and Simpson index. Result revealed that remittance contributed
highest to the household income followed by business and caste occupation, and rice farming.
Furthermore results showed that rural Bangladeshi households have diversified their livelihood
activities at medium level. The small and medium landholding households are more likely to
diversify their livelihoods compared to the functionally-landless and large landholding
households. The implication is that non-farm employment opportunities should be expanded to
combat poor households vulnerability to shocks and income fluctuations. It is also suggested
to give more attention to functionally-landless households for increasing and diversifying their
income.
Key Words: Livelihood diversification, simpson index, rural Bangladesh

INTRODUCTION
Bangladesh is an agrarian economy with 54% of its total
population residing in rural areas and involved directly or
indirectly in farming (World Bank, 2013).
Agriculture has been the primary source of livelihoods of
rural Bangladeshi since many years. But, recently
livelihoods are gradually diversifying away from
agriculture towards business, remittance, non-farm wage
labor,
agro-processing
and
cottage
industries,
construction and transportation operation, petty trade and
various services. Agriculture is no longer the key
occupation of rural livelihoods in Bangladesh (Hossain,
2010). Agricultures share to rural household income
dropped from about 60% in 1988 to 45% in 2008. One
major reason is the declining farm size. In Bangladesh,
the average landholding per household has declined from
0.61 ha in 1988 to 0.53 ha in 2000 and only 0.30 ha in
2013 (Hossain and Bayes, 2014). As a result, the role of

agriculture in rural livelihoods is also falling over time.

Besides, agriculture is a risky investment due to the
volatility in output price and weather risk. The impact of
risk and seasonality in agriculture triggered diversification
of rural occupations and income. On the other hand, nonfarm occupations reduce risk of income fluctuations by
combining activities that have different risk profiles, while
they can also ameliorate labor and consumption
smoothing problems associated with seasonality (Ellis,
2005). Through diversification, households find new ways
to raise and stabilize incomes thereby reduce
vulnerability to different livelihood shocks.
Corresponding Author: Md. Tanvir Ahmed, Department
of Agricultural Economics, University of the Philippines
Los Banos, Philippines. Email: tanvir4344@yahoo.com,
Tel.: +88029881265

Diversification of rural livelihoods in Bangladesh

Ahmed et al.

032

Diversified livelihood pattern is defined as combination of

various livelihood activities, such as crops, livestock, offfarm, and non-farm. Livelihood diversification can occur
by both agricultural diversification including producing
more crops per year or high-value crops; and nonagricultural diversification including migration, causal
labor, business, and services. Rural livelihood
diversification describes the phenomenon by which farm
households takes up non-farm activities, or rely on nonfarm income transfers to improve their standard of living
(Ellis, 2005). Livelihood diversification includes both farm
and non-farm activities which are undertaken to generate
income, additional to that of the main household activity,
via the production of agricultural and non-agricultural
goods and services, sale of waged labor, business or
self-employment in small firms, and other strategies to
minimize risk (Carter, 1997).
Literatures on rural livelihoods in Bangladesh are scarce.
Only limited studies have examined determinants of
household income in Bangladesh (Rahman, 2013;
Talukdar et al, 2013; Malek et al, 2009; Hossain, 2010).
But hardly any studies are available that examined the
livelihood diversification in rural Bangladesh. This study
assesses the patterns and quantifies the extent of
diversification of rural livelihoods.
Based on the primary data, this study provides empirical
evidence on how rural livelihoods are diversifying in
different regions of Bangladesh. It will help policymakers
and donors who frame policies and finance different
projects for the development of rural economy of
Bangladesh. The study will also provide methodological
guide for researchers in conducting studies related to
livelihood diversification.
DATA AND METHODOLOGY
Data Source and Sampling Design
The study was conducted in Bangladesh. A multi-stage
random sampling technique was used to select sample
villages. Eleven districts were purposively selected to
represent large geographical area and diverse livelihoods
of the country. In 10 districts, one sub-district from each
district, one union from each sub-district, and one village
from each union were selected randomly. In one largest
district in the country, two sub-districts, one union from
each sub-district, and one village from each union were
selected randomly. Thus, 12 villages were randomly
selected from 11 districts and four geographical regions
(northern, middle, south-eastern and western) of the
country. Finally, 45 rural households were randomly
chosen from each selected village making a total sample
of 540 households. Only 500 households were included
in the analysis as some households data were
incomplete.
The study used primary data collected through interview
schedule and pre-tested questionnaires; and the

information used were for 20122013 period. Information

elicited from the responded are demography, land
ownership, primary and secondary occupations of
household members, migrations and remittances, assets
ownership, labor force, on-farm activities, off-farm
activities, non-farm activities, credit and savings,
agricultural prices, income from different sources, and
living conditions.
The study used households net income by deducting
total cost from total returns. The share of income from
different sources was the basis to assess their livelihood
diversification. Extra attention was paid during data
collection and analysis to estimate households income
accurately because farmers do not keep record about
their crop production related data and often they tend to
under-report their income. Sometimes they do not
consider their own production and the in-kind received as
income.
Operational Definitions
The major variables used in the study are defined below.
Dependency ratio. It is the ratio of economically inactive
persons people younger than 18 or older than 59
over the active persons those ages 18-59 of a
household expressed in percentage.
Adult literacy rate. It represents the percentage of adult
people (aged >14 years) who can read and write.
Operated landholding. It is classified into four groups
based on the cultivated land: (1) functionally landless
(less than 0.2 hectare (ha)), (2) small (0.21-0.80 ha), (3)
medium (0.81-1.50 ha) and (4) large (over 1.50 ha).
Income source. Households income sources are
grouped in nine.
1)
Rice crop (net income from all rice crops in a
year),
2)
Non-rice crops (net income from all non-rice
crops in a year),
3)
Non-crop agriculture (income from livestock,
fishery and forestry),
4)
Agricultural
laborer
(labor
employed
in
agricultural sectors),
5)
Non-agricultural laborer (included both formal
and informal types of employment),
6)
Business and caste occupation,
7)
Salaried job and services,
8)
Remittance income (received from family
members
presently living outside the family: both
domestic and abroad), and
9)
Transfer payment.
Analytical Methods
Simple descriptive analysis (average, mean, median,
percentage etc.) along with the ANOVA test were carried
out to describe the socio-economic characteristics,
distribution of household income from different sources

Diversification of rural livelihoods in Bangladesh

J. Agric. Econ. Rural Devel.

033

Table 1. Demographic and socio-economic characteristics of the sampled households, Bangladesh

CHARACTERISTICS
Gender
Male headed household
Female Headed household
Age of HH Head
Age range of all members
0-14
15-64
65 > =
Dependency ratio
Household size
Male
Female
Adult literacy rate
Male
Female
Farm size
Land ownership class
Functionally landless
Small
Medium
Large
Land-Man ratio
Household Assets
Households with access to credit
Households who saved money in financial institutions

UNIT

(%)

%
%
(Years)

85.2
14.8

%
%
%
%
Number
Number
Number
%
%
%
(ha)

26.6
67.4
6.0
56.9

%
%
%
%
ha/person
BDT (USD)
%
%

28.2
43.0
19.6
9.2

MEAN

(+ STD.)

51.35

13.908

5.33
2.79
2.54

2.424
1.530
1.388

0.45

0.551

0.21
382,331 (4,780)

0.245
562073.9

83.8
88.7
78.9

43.4
57.2

Source: Survey result, 2012-13

and their share across the regions. Tabular analysis was

also employed to compare various household well-being
indicators of different study areas.
Livelihood Diversification Index
The most common measure of livelihood diversification is
the vector of income share associated with different
income sources (Khatun et al., 2012; Datta et al., 2011).
Besides, livelihood diversification is measured using
different indicators and indices, such as Simpson index,
Herfindahl index, Ogive index, Entropy index, Modified
Entropy index, and Composite Entropy index (Khatun et
al., 2012; Datta et al., 2011; Shaha et al., 2011; Shiyani
and Pandya, 1998). This study used Simpson index
because of its computational simplicity, robustness and
wider applicability. The formula for Simpson index (SID)
is:
2
SDI = 1- = i
Where, n is the total number of income sources and Pi is
the income proportion of i-th income source. The value of
SDI falls between 0 and 1. The indexs value is zero if
there is just one source of income. As the number of
sources increase, the shares (Pi) decline, as does the
sum of the squared shares, so that SID approaches to 1.
Households with most diversified income sources have
the largest SID value, and the least diversified income
sources have the smallest SID value. The higher the
number of income sources as well as more evenly
distributed the income shares, the higher the value of

SID. The Simpson index of diversity is affected both by

the number of income sources as well as by the
distribution of income among different sources. Based on
the SDI values, the level of livelihood diversification was
defined as:
1.
No diversification (SDI < = 0.01)
2.
Low level of diversification (SDI = 0.01 - 0.25)
3.
Medium level of diversification (SDI = 0.26 - 0.50)
4.
High level of diversification (SDI = 0.51 - 0.75)
5.
Very high level of diversification (SDI = > 0.75)
RESULTS AND DISCUSSION
Socio-Demographic Characteristics of Households
A demographic characteristic of any society is important
for analyzing its livelihood system. A society might be
poor in terms of endowment of natural resources but,
given
appropriate
policies
and
their
proper
implementation, a developed human resource could
show the paths to prosperity. In contrast, a resource-rich
society could pass through hardships in the presence of
rapid population growth and high dependency ratio
(Hossain, 2010). Table 1 shows the socio-demographic
characteristics of the selected households.
Out of the sample households, 85% were headed by
male and 15% were headed by female. This indicates
male dominant rural society of Bangladesh. The average
age of the household head was found to be about 51
years. Young family members prefer non-farm jobs in the

Diversification of rural livelihoods in Bangladesh

Ahmed et al.

034

Table 2. Distribution of household yearly income by sources

SOURCE OF INCOME

AMOUNT OF INCOME
(USD/YEAR)

Rice Crop
Non-rice Crops
Non-crop agriculture
Agricultural laborer
Non-agricultural laborer
Business and caste occupation
Salaried job and services
Remittances
Transfer payment
Total
F-value of ANOVA

380
202
221
61
171
487
155
699
16
2,393
14.27 (P = 0.000)

SHARE OF INCOME (%)

16
8
9
3
7
20
6
29
1
100

Source: Survey result, 2012-13, Note: 1 USD = 80 BDT

town or abroad and thus, the headship of the household

remains with the older member of the house. Among all
household members, about 67% was in the age group of
15-64 years, which indicates more economically active
persons in the house. High proportion of economically
active members is a human capital, which can generate
more income for the family. However, the estimated
average age dependency ratio of 57%, indicates that
there are 57 economically inactive member for each 100
economically active members. This implies that there are
still a considerable number of economically inactive
members present in rural household. The average family
size was 5.33 comprising 2.79 female and 2.54 male
members. The adult literacy rate was found 83.8%, which
is much higher as compared to the national adult literacy
rate of 57.7% (World FactBook, 2014).
The average size of operated land was small (0.45 ha).
This can hardly generate sufficient farm income for the
family of more than five members. The household must
find alternate source of income to maintain a basic living.
Of the total households, about 28% were functionally
landless, 43% had small farm size, 20% had medium
farm size, and only 9% had large farm size. The average
household assets value was computed to be about USD
4,800. About 43% of the total households had access to
credit while about 57% of them had access to savings.
Access to the financial dealings (credit and saving) by the
higher number of households indicates that those
households have more option to invest and to earn
additional income out of that investment.

share differ significantly across sources. Remittance

contributed highest share (29%) to the households total
income, followed by business and caste occupation
(20%) and rice crop (16%). The other sources also had
significant contribution.
The study also examined dominant income sources and
livelihood patterns across the four regions. In the
Northern region, rice crop and business and caste
occupation were the major sources of income
contributing 30% and 29.5% to the household income,
respectively (Table 3). The other important sources of
income were remittance (10%) followed by salaried job
and services (8%) and non-rice crops (7%).
Livelihood sources were different in the middle region.
Rice crop was the dominant (28%) source of income,
followed by remittance (25%). Besides, business and
caste occupation contributed 14%, non-crop agriculture
12%, non-agricultural labor wage 9% and salaried job
and services 7% of total household income.
In the South-Eastern region, remittance contributed
almost half (46%) of the household total income. The
second largest share (19%) earned from business and
cast occupation, followed by non-crop agriculture (10%).
The Western region had a balanced income share from
almost all sources. The maximum share (24%) in
household total income came from business and caste
occupation, followed by non-rice crops (22%). Nonagricultural laborer, salaried job and services,
remittances and rice crop accounted for 13%, 13%, 10%
and 7% of total yearly household income, respectively.

Sources of Household Income and Share

Household Livelihood Diversification
The most important determinant of livelihood for any
society is income. Income share from different sources
indicates the level of livelihood diversification. The
statistically significant F-value of ANOVA indicates that
livelihoods are highly diversified in rural Bangladesh. The
average annual total income of the sample household
was found to be about USD 2,400 (Table 2). Income

Several studies have used the Simpson index to measure

livelihood diversification (Shaha et al., 2010; Babatunde
et al., 2009; Joshi et al., 2003 and Hill, 1973).This study
followed the suite. Majority of the rural households
diversified their livelihoods into several activities and
earned significant amount of income from multiple

Diversification of rural livelihoods in Bangladesh

J. Agric. Econ. Rural Devel.

035

Table 3. The share of different sources in household yearly total income in four regions of Bangladesh.

SHARE OF INCOME (%)

SOURCES OF INCOME
Northern Region

Middle region

South-Eastern region

Western region

Rice crop

30.1

28.1

5.0

7.0

Non-rice crops

7.4

1.1

9.0

22.2

Non-crop agriculture

6.1

12.5

10.1

4.5

Agricultural laborer

1.0

2.0

2.6

5.5

Non-agricultural laborer

7.2

8.9

4.1

13.4

29.5

14.0

19.1

24.0

Salaried job and services

7.8

7.0

3.6

12.7

Remittances

9.7

25.5

45.9

10.4

Transfer payment

1.1

0.7

0.5

0.5

Total

100.0
98.3(P = 0.00)

100.0

100.0
8.64 (P = 0.00)

100.0
4.69 (P = 0.00)

Business and caste occupation

F-value

24.1(P = 0.00)

Source: Survey result, 2012-13

Table 4. Distribution of sampled household as per the level of livelihood diversification.

SDI RANGE
<=0.01
0.01 0.250
0.260 0.500
0.510 0.750
>= 0.760

PERCENTAGE (%)
6.0
19.6
31.8
38.4
4.2

LEVEL OF DIVERSIFICATION
No
Low
Medium
High
Very high

Source: Survey result and author computation, 2012-13

Table 5. Average value of Simpson diversification index across regions

REGIONS
Northern
Middle
South-Eastern
Western
All region (Cumulative)
F-value of ANOVA

NO. OF HOUSEHOLD
125
128
166
81
500

AVERAGE VALUE OF SDI

0.45
0.37
0.46
0.39
0.42

5.21 (P = 0.002)

Source: Survey result and author computation, 2012-13

sources. As depicted in the Table 4, 94% of the total

sampled
households
pursued
some
level
of
diversification in their livelihoods. Only 6% of households
had zero Simpson index, meaning they earned income
from just one source for their livelihoods. Of the sampled
households, 20% had low, 32% had medium, 38% had
high, and 4% had very high level of livelihood
diversification.
This implies that majority of the rural households are
diversifying their income.

Livelihood Diversification in Different Regions

The study examined if livelihoods differ across
geographical regions. The statistically significant F-value
of ANOVA indicates that livelihood diversification pattern
is different across regions. The cumulative average value
of the Simpson diversification index for all regions was
0.42, which means the households under the study fall
into medium level of diversification (Table 5). The table
also reveals that households in Northern, Middle, South-

Diversification of rural livelihoods in Bangladesh

Ahmed et al.

036
Table 6. Distribution of households level of livelihood
diversification and land holding size.

LAND CLASS

AVERAGE VALUE OF SDI

Functionally landless

0.34

Small

0.46

Medium

0.47

Large

0.40

All

0.42

F-value of ANOVA

10.11 (P = 0.000)

Source: Survey result and author computation, 2012-13

Eastern and Western region differed significantly on SDI

value. The highest average value (0.46) of Simpson
index was computed in South-Eastern region followed by
Northern region (0.45), Western region (0.39) and Middle
region (0.37).
Diversification of Livelihoods by Land Classification
The study examined if livelihood diversification vary by
landholding size. Nevertheless, functionally landless
households have limited scope to diversify their income
sources especially farm income compared to the small
and medium land holding households. Because, they are
always lacking of basic livelihood capitals (e.g. physical,
human, natural and financial) which pull back them to go
for livelihood diversification. Analysis found that,
functionally landless household had the lowest average
value of SDI (0.34) while the small and medium
household had highest average value of SDI (about
0.46). On the other hand large land holding households
may believe that they are secured against the
vulnerability to shock and risk. They always want to
augment their income by specialization rather
diversification. Thus, they dont go for higher level of
income diversification. The study found that average
value of SDI for large household was 0.40. There were
significant differences (at 1% level of significance) in the
average value of SDI among the four land classes of
households.

CONCLUSION
The Share of different sources to the household total
income differed significantly though income from
remittance, business and caste occupation, and rice crop
represented major source of rural livelihoods in
Bangladesh. Almost all households had some level of
livelihood diversification. About 32% of the total sampled
households diversified their livelihoods to the medium
and 38% diversified to the high level. The level
livelihood diversification varied significantly across the

geographical regions. The small and medium farmers

had more diversified livelihoods than the functionallylandless and large landholding farmers. Thus, it is
suggested to encourage the diversification of livelihood
activities by increasing and creating the opportunity of
more income generating activities especially for the
resource poor households in rural areas of Bangladesh to
combat the risk of vulnerability and income fluctuation.

ACKNOWLEDGEMENT
The author is very much grateful to the International Rice
Research Institute (IRRI) and Village Dynamics in South
Asia (VDSA) project for supporting this study by providing
necessary fund and data.

REFERENCES
Babatunde RO, Matin Q (2009). Patterns of income
diversification in rural Nigeria: determinates and
impacts. Quarterly Journal of International Agriculture.
48(4): 305-320.
Carter M (1997). Environment, Technology, and the
social Articulation of Risk in West African Agriculture.
Economic Development and Cultural Change. 45(3):
557-591.
Datta SK, Sing K (2011). Livelihood Diversification: Case
Study of Some Backward Region of India. International
Journal of Current Research, 3(2): 139-151.
Ellis F (2005). Small-Farms, Livelihood Diversification
and Rural-Urban Transitions: Strategic Issue in SubSaharan Africa. Paper presented for the research
workshop on The Future of Small Farms in Kent, UK.
Hill MO (1973). Diversity and Evenness: A Unifying
Notation and its Consequences, Ecology, 54(2): 427432.
Hossain M, Bayes A (2010). Rural Economy and
Livelihoods: Insights From Bangladesh. A H Development
Publishing House, Dhaka, Bangladesh.
Hossain M, Bayes A (2014). Rural Transformation in

Diversification of rural livelihoods in Bangladesh

J. Agric. Econ. Rural Devel.

Bangladesh, 1988-2013: Insights from Longitudinal

Surveys, BRAC seminar, BRAC Centre Inn, Dhaka,
Bangladesh.
Joshi PK, Gulati A, Birathal PS, Twari L (2003).
Agricultural Diversification in South Asia: Patterns,
Determinants and Policy Implications. Markets, Trade
and Institution Division Discussion Paper No. 57.
International
Food
Policy
Research
Institute.
Washington, D.C.
Khatun D, Roy BC (2012). Rural Livelihood
Diversification in West Bengal: Determinant and
Constraints. Agricultural Economics and Research
review, 25(1): 115-124.
Malek MA, Usami K (2009). Determinants of Non-farm
Income Diversification in Developed Villages of
Bangladesh. American Journal of Economics and
Business Administration , 1(2): 141-149.
Pederson G, Annou M (1999). Rural Household Income
Diversification in Vietnam. Quarterly Journal of
International Agriculture. 38: 2.
Rahman MS (2013). Socio-Economic Determinants of
off-farm Activity Participation in Bangladesh. Russian
Journal of Agricultural and Socio-Economic Sciences,
1: 13.
Shaha B, Bahal R (2010). Livelihood Diversification
Pursued by Farmers in West Bengal. Indian Research
Journal for Extension and Education. 10: 2.
Shiyani RL, Pandya HR (1998). Diversification of
agriculture in Gujrat: A spatio-temporal analysis. Indian
Journal Agricultural Economics, 53(4): 627-639.
Talukdar D, Chile L (2013). Determinants of Growth in
Consumption of Rural Household in Bangladesh: A
Regression Analysis. International Journal of Applied
Economics.
WORLD
BANK
(2013).
http://data.worldbank.org/indicator/SP.RUR.TOTL.ZS
WORLD FACTBOOK (2014). Population : Bangladesh:
The World Factbook (July 2014 est.)". Central
Intelligence Agency. Retrieved 2014-08-10.
Accepted 02 July, 2015
Citation: Ahmed MT, Bhandari H, Gordoncillo PU,
Quicoy CB, Carnaje GP (2015). Diversification of rural
livelihoods in Bangladesh. Journal of Agricultural
Economics and Rural Development, 2(2): 032-037.

Copyright: 2015. Ahmed et al. This is an open-access

article distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium,
provided the original author and source are cited.

Diversification of rural livelihoods in Bangladesh

037