CHAPTER OVERVIEW Survey Research Developmental Research Correlational Research.
Research Dholaikhal
description
Transcript of Research Dholaikhal
-
i
RESEARCH ON THE EFFECTIVENESS OF MICROINSURANCE IN BANGLADESH
Submitted to
Dr. Muhammad Ziaulhaq Mamun
Professor
Course: K301: Research Metholology
Institute of Business Administration
University of Dhaka
Submitted by
Mushreka Afroze Khan (RH-68)
Ornila Khan (RH-72)
Zeeshan Ahmed (ZR-82)
Waseem Khan (ZR-88)
Ishmam Ahmed Chowdhury (ZR-90)
Wais Al Karim (ZR-93)
Sudipta Saha Turja (ZR-95)
Hikmat Kabir (ZR-99)
Sanjir Ali (ZR-111)
Group: 1
Section: B, Batch:20th
Institute of Business Administration
University of Dhaka
July 6, 2014
-
ii
July 6, 2014
Dr. Muhammad Ziaulhaq Mamum
Professor
Institute of Business Administration
University of Dhaka
Sir:
Re: Submission of term paper
We would like to take this opportunity to present our term paper titled Research on the
Effectiveness of Micro Insurance in Bangladesh for your consideration as part of our Research
Methodology coursework. This report aims to address the effectiveness of Micro Insurance
schemes Bangladesh, including the history, issues, problems, challenges, and a survey of policy
holder to obtain an understanding whether the schemes are truly beneficial.
We have tried our level best to comply with your high standards and we sincerely hope that our
paper meets your expectations.
Sincerely
Mushreka Afroze Khan (RH 68) Ornila Khan (RH 72)
Zeeshan Ahmed (ZR 82)
Waseem Khan (ZR 88)
Ishmam Ahmed Chowdhury (ZR 90)
Wais Al Karim (ZR 93)
Sudipta Saha Turja (ZR 95)
Hikmat Kabir (ZR 99)
Sanjir Ali (ZR 111)
-
iii
ACKNOWLEDGMENTS
In course of our research work, being completely new to the field of insurance, let alone micro
insurance, the research team required industry help and information to guide this research
properly. For this we had reached out to two distinguished individuals from the insurance
industry, who had helped us, considerably and furthermore encouraged to look at their industry
at depth.
For their immense role in helping us obtain information and understanding about the insurance
industry we would like to convey our sincerest gratitude to the following people.
1. Mr. Munir Ahmed
Executive Director
Green Delta Insurance Co. Ltd.
2. Md. Manirul Islam
Managing Director & CEO
Pragati Insurance Limited
3. Sujit Kumar Bhoumik
Executive Vice President
Prime Insurance Company Limited
-
iv
TABLE OF CONTENTS
Title page
Letter of transmittal
Acknowledgements
Executive summary
1.0 INTRODUCTION
1.1 Issue
1.2 Problems
1.3 Objectives
1.3.1 Broad objectives
1.3.2 Specific objectives
1.4 Hypotheses
1.5 Scopes
1.6 Rationale
1.7 Limitations
2.0 LITERATURE REVIEW
2.1 The micro-insurance industry
2.2 Types of micro-insurance in Bangladesh
2.3 Problems in insurance industry
2.4 Insurance reform program
2.5 Micro-insurance for poverty alleviation
2.6 Who provides micro insurance
2.7 Micro-insurance and SME
2.8 Micro-insurance in public health
2.9 Micro-insurance as a promising sector
3.0 METHODOLOGY
3.1 Data collection
3.1.1 Primary source of data
3.1.2 Secondary source of data
3.2 Sample size
3.3 Sampling technique
3.4 Questionnaire development
3.4.1 Categories of questionnaire
3.4.2 Question patterns
3.4.3 Pretesting
3.5 Data collection
i
ii
iii
vii
1
1
1
2
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
9
9
9
9
10
11
11
11
12
12
13
-
v
3.6 Data Analysis
4.0 RESEARCH FINDINGS
4.1 Findings regarding knowledge of policyholders
4.1.1 Knowledge regarding number of companies
4.1.2 Knowledge regarding other programs of company the
respondent is a client of
4.1.3 Knowledge regarding the chosen micro-insurance policy
4.1.4 Knowledge regarding the reason for choosing the micro-
insurance policy
4.1.5 Synthesis of knowledge of policyholders
4.2 Findings regarding claims of micro-insurance
4.2.1 Findings regarding background check on claims
4.2.2 Findings regarding acceptance rate of claims
4.2.3 Findings regarding denial of claims
4.2.4 Findings regarding time of claims
4.2.5 Findings regarding satisfaction of policyholders relating to
claims
4.2.6 Synthesis of micro-insurance claims
4.3 Findings regarding collection and recovery of premium
4.3.1 Findings regarding medium of premium collection
4.3.2 Findings regarding considerations during premium collection
4.3.3 Findings regarding incidence of frauds
4.3.4 Findings regarding handling of frauds
4.3.5 Findings regarding time of recovery of premium after maturity
4.3.6 Findings regarding satisfaction about amount of premium
recovered
4.3.7 Synthesis of micro-insurance premiums
4.4 Findings regarding employees of micro-insurance
4.4.1 Findings regarding instances of misbehavior by employees
4.4.2 Findings regarding perception of employee behavior
4.4.3 Findings regarding the amount charged by agents
4.4.4 Findings regarding unrecorded payments
4.4.5 Findings regarding knowledge of employees
4.4.6 Findings regarding the frequency of visits made by agents
4.4.7 Synthesis of micro-insurance employees
4.5 Findings regarding coverage of micro-insurance
4.6 Combination of findings
5.0 CONCLUSION
13
14
14
14
15
15
16
17
18
18
20
21
21
22
23
24
24
25
26
27
28
29
29
30
30
31
32
32
33
35
35
36
38
39
-
vi
APPENDIX
Appendix 1: Questionnaires for the survey
Appendix 2:Co-ordination schema
Appendix 3: Estimated budget
Appendix 4: Time plan
Appendix 5: Hypothesis tests for the knowledge of policyholders
Appendix 6: Hypothesis tests for the claims of policyholders
Appendix 7: Bivariate correlation analysis
Appendix 8: Hypothesis tests for premium of micro-insurance
Appendix 9: Hypothesis tests for employees of micro-insurance
REFERENCES
41
41
45
47
48
49
51
55
56
60
65
-
vii
EXECUTIVE SUMMARY
Micro-insurance refers to insurance products designed for the low-income individuals. The word
micro represents the relatively small transaction size or lower premiums, a concept similar to
microfinance. Micro-insurance differs from traditional insurance in terms of the size of
premiums, coverage limits and target customers. The objectives of micro-insurance also vary
amongst different stakeholders. For governments and policymakers, for instance, micro-
insurance is a way to ensure inclusive growth and support the livelihoods of the vulnerable
segment of the society. For social and development organizations, micro-insurance can be an
effective tool to help alleviate poverty.
For a developing country like Bangladesh, micro-insurance seems like a boom that can help
Bangladesh address its problems regarding poverty especially in the agriculture and health
sectors. Indeed, micro-insurance has seen a lot of growth since its introduction in the country
during the 1970s with a growth rate of almost 15-20 percent each year since 2006, as said by Md.
Manirul Islam, Managing Director and CEO of Pragati Life Insurance Limited. Yet, a stellar idea
such as micro-insurance may not be a proper solution to Bangladeshs problems. In order to
justify the use of micro-insurance, we came up with a plan to gauge its effectiveness by setting
up 5 parameters that tests its relevancy here in Bangladesh. These parameters and their
effectiveness are as follows:
Knowledge of the policyholders of micro-insurance 25% effective based on 4 different
tests
Claims on micro-insurance 70% effective based on 5 different tests
Premium collection and recovery of micro-insurance 25% effective based on 6 different
tests
Employees of micro-insurance 60% effective based on 7 different tests
Micro-insurance coverage 10% effective based on 11 different tests
These effective rates were then combined and weighted as per the relevance of each parameter.
The final result showed that the total effectiveness of micro-insurance in Bangladesh stands at
only 34.2%, thus making it irrelevant as a development tool for Bangladesh. However, it should
be mentioned that the scope of this survey was limited to Dhaka only and may not be sufficient
enough to cover the overall scenario in Bangladesh. More extensive research needs to be carried
out in order to truly determine the usefulness of micro-insurance.
-
1
1.0 INTRODUCTION
1.1 ISSUE
Micro-insurance can be defined as an affordable subset of a financial service that uses risk
pooling to provide compensation to low income and poor individuals, entrepreneurs, households
or groups that are adversely affected by specific risks. While there may be great variance
between different micro-insurance schemes, they generally share a number of key characteristics
(Mizan R. Khan, 2013). These characteristics are:
a) Specifically targets low income and poor individuals and households
b) Designed to pool risks faced by the insured
c) Pricing is based on willingness to pay, and is proportional to the likelihood and cost of
the risks involved
d) Products are developed in collaboration with the communities they are supposed to
benefit
e) Products must be of substantive value to the poor in terms of addressing their
vulnerability to poverty
Micro-insurance was first introduced in Bangladesh in the 1970s in the form of health insurance
by an NGO called Ganashasthya Kendra. At present, there are over 60 micro-insurance
providers within the country and the growth rate of micro-insurance has been 33% between the
years 2008 and 2009. The sector is dominated by non-government organization microfinance
institutions (NGO-MFIs). Other types of organizations which operate in this sector are private
insurance companies and a couple of state-owned corporations (Mizan R. Khan, 2013).
1.2 PROBLEMS
Bangladesh is situated in the delta formed by multiple rivers as they meet the Indian Ocean
through the Bay of Bengal. Most of these rivers (Padma, Meghna and Jamuna for instance) are
originated from the Himalayas which lie to the north of the country. The lofty Himalayas in the
north and the funnel-like shape of the Bay of Bengal in the south have made Bangladesh one of
the worst victims of the catastrophic ravages of natural disasters like floods, cyclones, storm
surges, droughts, etc. These natural disasters render Bangladesh towards many losses: loss of
agricultural production, loss of livestock, loss of livelihood and loss of lives (Sakib, 2012).
-
2
Yet another notable fact about Bangladesh is that the country is highly dependent on its
agricultural sector. A total area of 14.943 million hectares are cropped within the country, may it
be singly, doubly or triply in a year. A staggering 47.5% of the total manpower of Bangladesh
earn their livelihoods from this sector and cumulatively account for 19.29% of the countrys
GDP as of 2013 (Ministry of Agriculture, 2014).
When both of the above are combined, the consequence that natural disasters have on the
economy of Bangladesh is overwhelming. Damages to the agricultural sector are defined as full
or partial destruction of assets in the sector. This includes destruction to agricultural land,
permanent plantations, irrigation or drainage systems, storage facilities, machineries, roads, etc.
Production losses occur due to the loss of a full crop due to the calamity, or as a result of a
decline in units yielded (Global Facility for Disaster Reduction and Recovery, 2008). When
methods are looked upon on how to tackle these production and financial losses which the
farmers suffer, one of the solutions that are often mentioned is the use of micro-insurance.
The span of micro-insurance is not only limited to the poor farmers from Bangladeshs
agricultural sector. One of the major applications of micro-insurance can be in the public health
sector. Bangladesh is a country where 88% of ones health expenditure comes from out-of-
pocket expenditure, 10% from non-profit institutions serving households and only a mere 0.8%
from pre-payment and risk pooling plans (Werner, 2009). This is fact enough to show what
micro-insurance is not very popular within Bangladesh.
This leads us to our research question. With so high growth in this sector and with so many
companies serving the general population, why is the micro-insurance not a common tool used
throughout Bangladesh? Is the micro-insurance sector effective?
1.3 OBJECTIVES
1.3.1 Broad objectives:
To find out the effectiveness of micro-insurance in Bangladesh
1.3.2 Specific objectives:
To find out the level of knowledge of the existing policyholders on the policies that are
being offered by the existing micro-insurance companies
To find out the coverage of micro-insurance within Bangladesh in terms of the assets that
are being insured by the existing micro-insurance policies
-
3
To find out the problems faced by the policyholders in the process of insurance claims
and settlements
To find out the problems faced by the policyholders in the process of payment of
premiums
To find out the problems faced by the policyholders in the process of recovering the
premiums following the maturity period
To find out the level of cooperation of the employees of the micro-insurance companies
1.4 HYPOTHESES
Some hypotheses have been derived out of the specific objectives based on our expectations:
The existing policyholders do not have much overall knowledge on micro-insurance.
The claiming process is not effective in the context of micro-insurance in Bangladesh
The process of collection of premium is effective, but that of recovery of premium after
maturity is not.
The coverage of micro-insurance in terms of assets is quite high in Bangladesh.
1.5 SCOPES
The research topic is only limited to micro-insurance. All the other types of insurance
policies are not within the scope of the report.
The paper only considers the opinions of surveyed policyholders from different parts of
Dhaka city and a few selected companies which provide micro-insurance policies.
For all the hypothesis testing that has been carried out in the research process, the
significance level () that has been taken into consideration for drawing a conclusion is
5%.
-
4
1.6 RATIONALE
This report was assigned as a compulsory content in Research Methodology course by our course
instructor. Alongside, there are several beneficiaries of this report.
Companies providing micro-insurance service
Insurance Development and Regulatory Authority Bangladesh (IDRA)
Institute of Microfinance (InM)
Research institutes and individual researchers
Existing and potential policyholders of micro-insurance
1.7 LIMITATIONS
The research area is only limited to the district of Dhaka, the capital city of Bangladesh.
Due to shortage of resources, other districts were not taken into account. So all the
samples taken are from the Dhaka city.
The biggest limitation faced in the process of the research was due to the fact that no self
administered questionnaire was valid. Interviews needed to be carried out personally by
the group mates, which decreased the span of the sample size covered.
Some of the respondents were reluctant to answer to some of the questions in the
questionnaires. Hence, those values were entered into the database as missing values.
Due to unavailability of data regarding some of the aspects of the research paper, some
approximate values were used in certain areas of the research.
-
5
2.0 LITERATURE REVIEW
2.1 THE MICRO INSURANCE INDUSTRY
In 1980, Grameen Bank of Bangladesh started lending to poor people without collateral security,
thereby revolutionizing finance and banking. Inspired by this scheme, a life insurance company
Green Delta of Bangladesh came forward in 1988 to provide financial security to the poor in
the form of micro insurance at a small amount of monthly premium.
After its introduction in Bangladesh, micro insurance spread among other life insurance
companies at a rapid pace. To date, almost every life insurance company in Bangladesh operates
at least more than one micro insurance project. In 2008, about two million new policies were
sold under micro insurance, compared to a million under ordinary individual life. The total micro
insurance premiums in 2008 amounted to approximately half of ordinary and other life
premiums.
For the last few years, micro-insurance portfolios of different companies have grown at an
average rate of more than 20% per annum (Ali M. K., 2009). This spectacular growth of micro-
insurance in such a short period reflects the necessity and acceptability of micro-insurance
among the masses in the country. Given that this trend is expected to continue in the years to
come, premium income under the micro-insurance portfolio will likely to overtake ordinary life
premium.
Currently the insurance industry has 77 insurance companies in Bangladesh, 75 private insurance
companies and 2 public insurance companies (Insurance Development and Regulatory Authority
Bangladesh, 2014).
The insurance sector in Bangladesh is relatively small. However, the sector has shown
remarkable growth in recent years. Parliament on 3rd
March 2010 passed two insurance laws in a
bid to strengthen the regulatory framework and make the industry operationally vibrant. The new
laws, came into effect on 18 March 2010, are Insurance Act 2010 and IDRA 2010 (Insurance
Development and Regulatory Authority Bangladesh, 2014).
While the Insurance Act 2010 has been framed with the view to synchronize functions of the
existing Insurance Department, the IDRA 2010 aims to make an independent regulatory body
that will overlook the industry and protect customers interests. The Insurance Act 2010 now
recognizes and brings under it other than the normal insurances, Islami insurances and micro
insurance businesses.
The act is making way for the micro insurance business opportunities in the insurance sector of
Bangladesh, which has a great prospect for small and medium enterprises as well as the growing
businesses, especially in the rural areas.
-
6
2.2 TYPES OF MICRO INSURANCE IN BANGLADESH
1. General Micro Insurance
This type of micro insurance aims to provide security towards everything other than life of the
underprivileged in the economy. As a normal general insurance applies, the micro general
insurance provides the same type of security at a much lower premium. As observed through the
survey, the team recognized that this type of insurance is rare in the micro insurance scenario.
Only recently has it been started to be of use by SMEs (Small Medium Enterprises) due to
insurance being a condition for micro finance. The insurance is still limited to events of fire.
2. Micro Life Insurance
Like normal life insurances, micro life insurances provide security against the policy holders
life. The premiums are lowered to make it available for the lower income group segment of the
country. From the survey observations, this type of micro insurance is yet to be popular with the
target market.
3. Microtakaful (Islami Insurance)
This segment of micro insurance is the type of micro insurance being provided by Islami
insurance companies.
4. Deposit Pension Schemes (DPS)
This is the most popular micro insurance plan launched by the commercial life insurers. This
gained popularity not because of the benefits it provides in relation to its premium rate but
mainly because of the familiarity among the common people of a deposit scheme introduced by
some of the commercial banks with the same title. The premium rates of DPS is determined
usually by dividing the amount of sum assured with the number of premium installments payable
during the term of the insurance. A maturity, total amount of premiums paid during the term is
payable together with the accrued bonuses. A policyholder, instead of receiving the total amount
of premiums at maturity, may exercise the option of pension benefit.
2.3 PROBLEMS IN INSURANCE INDUSTRY
In Bangladesh, the term insurance is not very common. It is limited to only a certain
proportion of the urban community and almost non-existent within the rural community. This
problem arises because of a number of factors as stated by a research done on ALICO,
Bangladesh. The first and foremost of them is the fear of the public of being cheated. People
have preconceived ideas about the existing insurance policies and do not welcome the new ideas
as much as the company hopes they would. Moreover, all the documents of ALICO (and most of
-
7
the private insurance companies) are in English rather than Bengali. On the other hand, ALICO
faces an array of problems too in Bangladesh (some of which are common to other insurance
companies). The rules and regulations regarding this sector are very strict in Bangladesh. The
company itself is afraid of cheats who provide wrong information to collect claims (some cheats
have gone even to the extent of opening fraud branches). The article states that lack of planning,
lack of capital and lack of motivation are three major reasons for which the insurance industry
has not had its desirable influence over the country. (The Lawyers and Jurists, 2013)
2.4 INSURANCE REFORM PROGRAM
The government of Bangladesh is trying to improve the insurance acceptance by embarking on a
reform program. As a first step, the government has replaced the Insurance Act, 1938 with the
Insurance Act, 2010. One of the barriers for starting an insurance company is in the
unavailability of the capital. None of the private life insurance companies have been able to raise
a capital of more than Tk. 30 crore whereas the ideal amount should have been greater than
twice. A main lacking that has been pointed out is the lack of proper training in Bangladesh
Insurance Academy (BIA) and Bangladesh Institute of Bank Management (BIBM). (Ahmed,
2011)
2.5 MICRO INSURANCE FOR POVERTY ALLEVIATION
Micro-insurance is a concept which bridges the two terms micro and insurance to provide a
solution for this insurance deprived nation. Micro insurance refers to the protection of low
income people against specific perils in exchange for regular premium payment proportionate to
the likelihood and cost of the risks available. This way, the target segment becomes more
complete: the 40% of the population of Bangladesh who are stricken by major poverty. The
major cause of this poverty is the deprivation of land and non land asset, access to education,
remunerative occupation and opportunities of diversified income sources. If productivity is
increased, it will alleviate the poverty to a significant level. To aid this fact, many insurance
companies have brought out a single comprehensive micro insurance package to cover for both
life insurance and property risks. What it ensures is that the possibility that some peril may
interrupt the income of the poor household is greatly reduced. The areas in which micro-
insurance has a prospect are mainly: dwelling, stables, stores and shops; pump sets, harvesters,
threshers; handicrafts and household productions; and personal accident and hospitalization.
Some MFI-NGOs are now offering the service of micro-insurance in the name of self
insurance. (Ali K. M., 2011)
-
8
2.6 WHO PROVIDES MICRO INSURANCE
Micro insurance providers can be broadly divided into two categories:
a) Mainstream insurance companies (example: Delta Life Insurance Company)
b) NGOs / MFIs (example: Grameen Kalyan, Sajida Foundation, Shakti Foundation, etc.)
Among the NGO/MFIs some specialize in the provision of healthcare (example: GK), while
others provide a range of financial services as well (example: BRAC and Shakti
Foundation).(Syed M. Ahsan, 2010)
2.7 MICRO INSURANCE AND SME
In Bangladesh, there exist a huge number of small and medium enterprises. The definition of
SMEs, though poses many controversies, consistently point out that the businesses contain a
small number of employees (below 25 as per Bangladesh Bank) and a small spread of fixed
assets other than land and building (below Tk. 5,000,000 as per Bangladesh Bank) (Bangladesh
Bank, 2010).
In many cases, the assets held by these SMEs are insured by insurance firms. As per a South
African case study, it was stated that low-income entrepreneurs are particularly vulnerable to
risks. Lacking adequate financial and other assets, the poor are exposed to the smallest economic
shocks. Their vulnerability is exacerbated by the fact that low-income persons tend to live and
work in riskier environments than wealthier people, with a greater likelihood of illnesses,
accidents and thefts (Aliber, 2001).
This is also very true in case of Bangladesh. The case study also pointed out that as the family of
most of the entrepreneurs depends highly on income from this business, the insurance policies
specifically aimed towards them can also be categorized as micro-insurance.
2.8 MICRO INSURANCE IN PUBLIC HEALTH
One major prospect of micro insurance is the public health sector. It has been stated that micro-
insurance can lower both the ongoing, preventive health cost and also the high catastrophic
health cost for families affected by poverty in Bangladesh. The micro-insurance scheme of India
has been criticized and it has been stated that that scheme will not apply to Bangladesh. This
micro-insurance will reduce the tendency of rural people to turn to minimally-trained village
pharmacists or traditional healers. Instead they will seek professional healthcare due to risk
pooling. But to achieve all these, the initiatives must exercise adept level of professional
-
9
management, product development, management information system and re-insurance (Werner,
2009).
2.9 MICRO INSURANCE AS A PROMISING SECTOR
The concept of micro insurance stems from Dr. Mohammad Yunuss idea of micro credit lending
which initially began in 1980s. The first insurance company to offer micro insurance was Green
Delta of Bangladesh. Since then, micro insurance in Bangladesh has come a long way and
currently it constitutes on average 30% of the income generated. Over the last few years, micro
insurance growth has been at around 20% on average, which is quite commendable. In the
formal side, the most popular scheme is the Deposit Pension Scheme (DPS), which gained
popularity not because of its competitive premium rates, but because of peoples familiarity with
the term DPS as it was introduced earlier by some commercial banks. During early 2000, the
Government of Bangladesh started permitting Islamic Insurance companies; their insurance plans
are known as Takaful. The informal sector of micro insurance is dominated by NGOs. Almost
all major NGOs have micro insurance schemes which cover various factors such as outstanding
loan balance, health, disability and in some cases even provide one time monetary benefit. In
many cases, these micro insurance schemes are financed partly by donor funds. One of the major
challenges faced by micro insurance is the lack of proper regulation. While there are clear
guidelines for general insurance practices, rules and regulations are not as well put for micro
insurance plans. For a healthy growth and consistent growth in the micro insurance segment, a
proper set of rules and regulations must be set up and implemented (Ali M. K., 2009).
3.0 METHODOLOGY
3.1 DATA COLLECTION
3.1.1 Primary source of data
Existing policyholders of different micro-insurance policies have been interviewed
located within the area of Dhaka city.
Micro-insurance policymakers have been interviewed to understand the behavioral
pattern of the policyholders and how it affects the policy.
3.1.2 Secondary source of data
It includes various journals and reports on the Micro-Insurance sector of Bangladesh. By
reviewing journals and reports on the micro-insurance sector of Bangladesh, we can understand
-
10
the key factors that are posing barriers to the effectiveness of the policies. Past reports also
present to us the dimensions that have already been ventured into by researchers, allowing us to
save time and money and the opportunity to look into new ventures ourselves.
All of the secondary sources of data used in the research process have been mentioned in the
Reference section of the report.
3.2 SAMPLE SIZE
Currently, micro-insurance covers 6.5 million lives in Bangladesh (Werner, 2009). So, this
population size can be used to determine the number of existing policyholders that we need to
survey. In this case, the targeted sample size can be determined with the following formula:
Here,
n = Sample size
N = Population size
p = Proportion of the population who are micro-insurance policyholders
q = Proportion of the population who are not micro-insurance policyholders
z = Reliability, depending on the level of significance
do = Precision
We assume the product of p and q to be 0.25 (the maximum possible values). For a level of
significance of 5%, the value of z is 1.96. We assume a precision, do of 5%. In that case, the
sample size is:
However due to time and resource limitation, the actual sample size that could be considered for
the whole research process has been 40. For this sample size and using the same formula, the
precision of the research carried out has been calculated below.
-
11
So, Precision = do = 15.50%
3.3 SAMPLING TECHNIQUE
For the purpose of this study, non-probability sampling was used.
Non-probability sampling represents a group of sampling techniques that help researchers to
select units from a population that they are interested in studying. The main purpose for selecting
the non-probability sampling is the limited time and workforce, lack of access to the entire target
population and practicality (easier, quicker and cheaper).
We have taken a combination of judgmental and snowball approaches in coming up with the
desired sample of the policyholders. We have employed the judgmental approach in selecting the
areas from which surveys will be taken and in selecting the companies that were interviewed
during the research process. We have employed the snowball approach in selecting individual
policyholders, since information about them is not available to public at large.
3.4 QUESTIONNAIRE DEVELOPMENT
After finalizing the co-ordination schema to etch out the variables that will be used in the
research process [Appendix 2], questionnaires were designed to carry out the survey [Appendix
1].
3.4.1 Categories of questionnaire
To reach out to the interviewees, two questionnaires were formulated:
a) Questionnaire for the existing policyholders of micro-insurance policies:
This is the main questionnaire for the research since most of the variables listed in the co-
ordination schema are aimed at the existing policyholders of micro-insurance.
b) Questionnaire for the micro-insurance policymakers:
To cross check some of the complex variables and to check for the coverage of micro-insurance,
interviews were carried out to some of the companies providing micro-insurance services. The
names of the personnel who were interviewed are present in the Acknowledgements.
-
12
For all the two questionnaires formulated, the respondents will be interviewed since many of the
policyholders may not know how to read or write. For their benefits, the first questionnaires
addressed towards the existing policyholders of micro-insurance will be translated into Bengali
so that the respondents understand the questions better. The interview of the micro-insurance
policymaking companies will be carried out in English.
3.4.2 Question patterns
For the questionnaire aimed towards the existing policyholders of micro-insurance, the following
answering bases were used:
a) Subjective questions
b) Yes/No questions: Here the answer of Yes has been given a value of 1 and the answer
of No has been given a value of 0
c) Checklist of items
d) 5 point Likert scale: The value of the question patterns range from -2 to +2
For the interview questionnaire aimed towards the policymakers, subjective questions were
asked to gather relevant information.
3.4.3 Pre-testing
The questionnaire aimed towards the existing policymakers was pre-tested among 7 respondents.
On carrying out the pre-test, the following changes were made:
a) For question number C1, the first option, Cash upfront was divided into two separate
options of Through office and Through agents.
b) For question number C5, the previously subjective question was changed to include the
options Fraud agents, Fraud companies, Fraud paperwork and Others for the
reason of specificity.
3.5 DATA COLLECTION:
The data collection spanned over a period of 4 weeks. During this time, several areas were
covered up in an aim to carry out interviews with probable respondents. The major limitation
faced in this process was that the group had to take interviews of the respondents instead of
handing them a self administered questionnaire.
The areas that were taken into consideration are listed below.
-
13
Dhaka University
Kataban
Nilkhet
Uttara
Eskaton
Shahbagh
Dhanmondi
3.6 DATA ANALYSIS
For the purpose of data analysis, the collected data were entered into SPSS and different
statistical tests were administered. The tests that were carried out include:
a) One sample T-Test: Hypothesis tests were carried out for different variables where the
sample means of the data collected for those variables were used to check whether the
population means of those variables conformed to different standard values.
b) Frequency analysis: The frequency of answers as per the choices to different questions
was presented where necessary and the results were explained as to how it can be related
to the effectiveness of micro-insurance.
c) Correlation coefficients: For some variables, the correlation has been found out to see the
strength and direction in which they move.
-
14
4.0 RESEARCH FINDINGS
4.1 FINDINGS REGARDING KNOWLEDGE OF POLICYHOLDERS
4.1.1 Knowledge regarding number of companies
As already stated in the preceding sections of the report, the total number of insurance companies
in Bangladesh is 77 (Insurance Development and Regulatory Authority Bangladesh, 2014) of
which as high as 60 companies provide micro-insurance facilities (Mizan R. Khan, 2013).
Moreover, majority of the respondents are policyholders of micro life insurance. There are in
total 30 companies which provide life insurance (Insurance Development and Regulatory
Authority Bangladesh, 2014). Thus, for the hypothesis test carried out below, it has been
assumed that all of these life insurance companies provide micro-life insurance.
In this section, the claim that the policyholders do not know the correct number of companies has
been tested statistically. It has been assumed, as stated above, that the correct number of
companies is 30. Thus, the correct number of companies other than the company of which the
respondent is a client of is 29.
The set of hypotheses taken into consideration and the tables generated by SPSS are given in
[Appendix 5], under the subtopic of Hypothesis 1. A simpler table showing the mean of the
sample and the significance level has been given below.
Sample mean ( ) 4.56
Significance level for 2 tailed test 0%
The mean perceived value of the sample (x ) in this context is only 4.56 which shows, to a very
large extent, the respondents are unaware of the correct number of companies.
The significance level of the two tailed test carried out shows that the value is 0% (as underlined
in the table above). This means that at any significance level above 0%, we can reject the null
hypothesis that the mean number of companies perceived by policyholders is 29. The alternate
hypothesis that the mean number of companies perceived by the policyholders is not 29 is
accepted.
Conclusion: At a significance level of 5%, the policyholders of micro-insurance do not have the
appropriate knowledge regarding the number of companies currently in the industry.
-
15
4.1.2 Knowledge regarding other programs of company the respondent is a client of
In this section, we will check whether the respondents think that they have good idea about the
other insurance offers of the company that he/she is a policyholder of. One assumption that is
made here will be that even if half the respondents feel that they have a good understanding of
the other programs being offered by the insurance company that he/she is a policyholder of, we
can conclude that the micro-insurance companies are effective in this context.
The set of hypotheses taken into consideration and the detailed tables generated by SPSS are
given in [Appendix 5], under the subtopic of Hypothesis 2. A simpler table showing the
frequency distribution of the answers and the significance level has been given below.
Knowledgeable Unknowledgeable
Frequency distribution 31 9
Proportion of frequency 22.5% 77.5%
Significance level for 2 tailed test 0%
From the tables it can be seen that the sample mean for this variable as received from the
questionnaire is 0.225. This means that only 22.5% of the respondents claimed that they had
knowledge regarding the other programs of the same insurance company.
The significance level for a two tailed test is 0% in this context. As this hypothesis testing had
been a one tail test, the significance level for two tail test is to be divided by 2 to get that of the
one tail test. Even after doing that, the significance level is 0%.
Thus, it can be concluded that for any significance level greater than 0%, the null hypothesis that
at least half the policyholders have a good about the other programs of the company that they
are the policyholder of is rejected and the alternate hypothesis that less than half the
policyholders believe they have good idea about the other programs of the company that they
are the policyholder of is established.
Conclusion: At a significance level of 5%, a proportion of less than 50% of the policyholders of
micro-insurance have the appropriate knowledge regarding the other micro-insurance programs
that the company that he/she is a client of offers.
4.1.3 Knowledge regarding the chosen micro-insurance policy
In this section, we will check whether the respondents think that they have a good understanding
of the micro-insurance program that they are currently under. In this case too, the assumption
that if half the respondents feel that he/she has good knowledge regarding the chosen program,
the micro-insurance companies can be tagged as being effective in this context.
-
16
The set of hypotheses to be considered for this part and the tables generated by SPSS are
included in [Appendix 5] under the sub heading of Hypothesis 3. A simpler table showing the
frequency distribution of the answers and the significance level has been given below.
Knowledgeable Unknowledgeable
Frequency distribution 24 16
Proportion of frequency 0.6 0.4
Significance level for 2 tailed test 21%
From the above tables, we can see that the sample mean for this category is 0.60. This means that
60% of the respondents believe that they have good understanding of the micro-insurance policy
that he/she has taken up.
The hypothesis test carried out gives its result in the next table. The significance level being
shown in this context is 21% for a two tailed test. However, this is a one tail test for which this
level of significance must be divided by 2. Thus, the significance level for the one tail test is
10.5%.
So it can be said that for any value of significance level below 10.5%, the null hypothesis that at
most half the policyholders believe they have a good understanding of the chosen insurance
program cannot be rejected and the alternate hypothesis that more than half the policyholders
believe they have a good understanding of the chosen insurance program cannot be established.
For any value of significance level above 10.5%, the null hypothesis that at most half the
policyholders believe they have a good understanding of the chosen insurance program is
rejected and the alternate hypothesis that more than half the policyholders believe they have a
good understanding of the chosen insurance program can be established.
Conclusion: At a significance level of 5%, it cannot be concluded that a proportion more than
50% of the policyholders of micro-insurance have appropriate knowledge regarding the chosen
micro-insurance policy.
4.1.4 Knowledge regarding the reason for choosing the micro-insurance policy
In this section, the answers of the question to A5 [Appendix 1] has been arranged into two
groups: whether the respondents have taken the policy knowing its repercussions or whether they
have taken the policy by being affected by the words of another person or party.
A hypothesis test has been carried out below to check the claim whether statistically the
proportion of the population aware of the repercussions of the policy is greater than 50%. The set
-
17
of hypotheses to be considered for this part and the tables generated by SPSS are included in
[Appendix 5] under the sub heading of Hypothesis 4.
A simpler frequency table and the values of the significance level as per the hypothesis test has
been given in the table below.
Knowledge of the
repercussions
Influence of other
party
Frequency 28 12
Proportion p = 0.70 q = 0.30
Significance level for 2 tailed test 1%
From the above table, we can see that the significance level for two tail test is 1%. But as the
hypothesis in this context is a one tail test, this percentage must be divided by 2. So the
significance level for one tail test is 0.5%.
Thus, it can be said that for a level of significance less than 0.5%, the null hypothesis that no
more than 50% of the policyholders know the reason for choosing the policy that they are under
cannot be rejected and the alternate hypothesis more than 50% of the policyholders know the
reason for choosing the policy that they are under cannot be established. For a level of
significance greater than 0.5%, the null hypothesis can be rejected and the alternate hypothesis
that more than 50% of the policyholders know the reason for choosing the policy that they are
under can be established.
Conclusion: At a level of significance of 5%, it can be concluded that the more than half the
policyholders of micro-insurance know the reason for choosing the micro-insurance policy.
4.1.5 Synthesis of knowledge of policyholders
To combine the results from all the simple variables into the parameters, the conclusions from
the individual simple variables have been rated either 1 or 0. A value of 1 means that in
terms of that simple variable, the micro-insurance in Bangladesh.
For the knowledge of the policyholders, the simple variables and their corresponding values as
per the conclusions are listed in the table below:
Simple variable Conclusion Value
Knowledge regarding
number of companies
The policyholders do not have appropriate
knowledge regarding the number of companies.
0
-
18
Knowledge regarding other
programs of the company
Less than half the policyholders have appropriate
knowledge regarding other policies of the company.
0
Knowledge regarding the
chosen micro-insurance
policy
More than half the policyholders do not know the
reason regarding the choice of the micro-insurance
policy.
0
Knowledge regarding the
reason for choosing it
More than half the policyholders know the reason
for choosing the policy.
1
Average value 0.25
4.2 FINDINGS REGARDING CLAIMS OF MICRO-INSURANCE
4.2.1 Findings regarding background check on claims
In the Insurance industry claims turn out to be a major hurdle for the policyholders to be
applying for claims. The first stage of the claim process in micro insurance is for the policy
maker to validate the claim of the policyholder by performing a thorough background check via
independent third party.
In the case of our sample, we first tested for how many of the policyholders within the sample
have actually claimed for insurance and secondly we tested how thorough the background check
was according to the policyholders who have claimed their insurance.
In the aspect of how many insurance claims, we made a frequency table as shown below. The
table helps us understand how many of our sample has actually made a claim. Within the sample
only 7 policyholders have made a claim in the tenure of their policy, while 32 policyholders have
never applied for an insurance claim.
Made an Insurance Claim
Frequency Percentage
Yes 7 17.5
No 32 80.0
Total 39 97.5
Mean .18
-
19
Significance Value (two tailed) .006
In the first part of this section we will check as to whether any of our samples have made a claim
on their insurance. An assumption here is that the sample policyholders have not yet made a
claim against their insurance scheme.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 6], under the subtopic Hypothesis 1. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained above.
We can observe from the table that the sample mean for this variable as received from the
questionnaire is 0.18. This means that only 18% of the sample has made claims against their
insurance scheme.
The significance for a one tailed test is 0.06% for this simple variable. In case of a two tailed test,
the significance value would have been divided by 2 to get that of a one tailed test, and obtain
0.03%.
Hence it can be concluded that for any significance level greater than 0.03% the null hypothesis
that a significant proportion of the policyholders have not made claims is rejected and the
alternate hypothesis a significant proportion of the policyholders have made claims is
established.
In the second part of this segment we try to find out whether the claims that have been made
within the sample were validated with proper background check. The assumption we have made
for this part is that none of the claims that were made, were validated with a thorough
background check.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 6], under the subtopic Hypothesis 2. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
Mean Significance Value
Thorough Background Check 0.78 0.00
From the table it can be seen that the sample mean for the simple variable as received from the
questionnaire is 0.78. The significance level for the two tailed test is 0 in this context. If we turn
it into a one tailed test and divide the significance value with 2, we would still obtain a value of
zero.
-
20
Thus it can be concluded for any significance level greater 0%, the null hypothesis that the
policyholders perceive that the background check on claims is not thorough is rejected and the
alternate hypothesis that the policyholders perceive that the background check on claims is
thorough is established.
Conclusion:
At significance level of 5%, a significantly large number of policyholder has made a
claim against their insurance scheme.
At a significance level of 5%, the policyholders of micro-insurance perceive that the
background check on claims is thorough.
4.2.2 Findings regarding acceptance rate of claims
In this segment we will test to see whether the claims that have been made by the respondents
were accepted by the policymaker. Here again we assume that the acceptance rate of the claims
made by the respondents is equal to or less than 15. 15% is the industry standard rate of
acceptance. For which we test whether the respondents of our sample have a different rate or not.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 6], under the subtopic Hypothesis 3. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
Frequency Mean Significance Value
Acceptance Rate of
Claims
5 0.48 0.000
We can see from the table that the sample mean for the simple variable is 0.48 based on the
questionnaires. This means that 48% of the respondents have experienced their claim to be
accepted by the policymaker.
The significance level for this simple variable is 0.000 in a two tailed test. Considering a one
tailed test, we divide the significance value by 2 and still obtain 0.000. We would still have a
significance level of 0%.
We can come to the conclusion that for any significance level greater than 0.0%, we reject the
null hypothesis that acceptance rate of claims is equal to or less than zero and the alternate
hypothesis acceptance rate of claims is greater than zero is established.
Conclusion: With a significance level of 5%, majority of the respondents say that their claims
have been accepted by the policy makers.
-
21
4.2.3 Findings regarding denial of claims
In this section we will observe as to why there was any denial of claims amongst the
respondents.
Frequency Percentage
Claims Made 7 100
Claims Denied 2 28.6
Reasons: 1. Claim before maturity not provided
2. Policy nullified.
As you can see in the table above, a total of 7 insurance claims were made within our sample,
while 5 of the claims were accepted, only 2 were rejected. As the reason show, one of the claims
were rejected because of the firms policy of not providing claim before maturity and another
being that the policy was made void.
If we look at the percentages we can see that 28.6% of the claims made were rejected. Although
the percentage appears to be high, we must remember that 100% means only 7 claims. So the
percentage can be said to be exaggerated.
Conclusion: The information obtained is not enough for us to measure the effectiveness in the
denial of claims of policyholders. Value of Effectiveness: 0.5
4.2.4 Findings regarding time of claims
In accordance to question number B6 [Appendix 1], this section looks into how much time is
taken for the policymakers to pay off the premium of a claim.
In the sample only four respondents have received their claim amount during the survey. Out of
five of the claims which were accepted, four claims were paid off within a maximum of 3
months and a minimum time of zero. The variation cant be used as an average amongst the
micro insurance industry as the number of claims received is low.
Time Taken for Claim to be Given
(in months)
Frequency Percentage
0 1 0.25
1 1 0.25
2 1 0.25
-
22
3 1 0.25
Total 4 100
Conclusion: Although we can say that 4/5 or 80% of the claims were paid off, however we cant
say whether this aspect of Insurance Claim is effective or not because there are only four
samples, causing each time taken to have an equal percentage of 25. Value of Effectiveness: 0.5
4.2.5 Findings regarding satisfaction of policyholders relating to claims
We divided this section of the findings into two parts. The satisfaction of the policyholders is
firstly tested with satisfaction with the time taken for the company to evaluate the claim and
respond and the time taken for the company to provide the claimed amount. Secondly we test
and find out the satisfaction of policyholders with the claim amount they have obtained.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 6], under the subtopic Hypothesis 4 & 5. A simpler table of showing the
frequency distribution of the answers, the mean value and the significance value obtained are
given accordingly.
Mean Significance Value
Satisfaction with B4 and B6 1.00 0.089
B4 = Satisfaction with time taken for claim evaluation and response.
B5 = Satisfaction with time taken for the claim amount to be given.
We can see that the sample mean for the first satisfaction related variable is 1.00.The
significance value for the satisfaction variable is 8.9% in a two tailed test. Considering a one
tailed test for this simple variable, we obtain the significance level of 4.45%.
Hence we can come to the conclusion that any significance level greater than 4.45% would mean
that we reject the null hypothesis of policyholders are not satisfied with the time taken to
evaluate and provide the claim amount and establish that claim that they are indeed satisfied
with the processes.
The second part of this segment is whether the respondents are happy with the premium amount
they receive after a claim is made. The claim for this part is that the policyholders are satisfied
with the claim amount that they receive.
-
23
Mean Significance Value
Satisfaction the Amount of Claim Received 0.40 0.587
From the table we can see that the sample mean for the variable is 0.40 in this context. We can
also see that the significance level for the variable is 58.7%. Thus the significance level for a one
tail test is 29.4%.
Finally we can come to the conclusion that any significance level greater than 29.4% would
result in rejection of the null hypothesis, the policyholders are not satisfied with the claim
amount, and while establishing the alternate hypothesis that they are satisfied. For a
significance level lower than 29.4%, the null hypothesis cannot be rejected and the claim cannot
be established.
Conclusion:
With a significance level of 5%, we can say that the policyholders are satisfied with the
time taken for claim evaluation and time taken for the claim to be received.
For the significance level of 5%, we must accept the null hypothesis that respondents are
not satisfied or neutral with the claim amount they receive.
4.2.6 Synthesis of micro-insurance claims
To combine the results from all the simple variables into the parameters, the conclusions from
the individual simple variables have been rated either 1 or 0. A value of 1 means that in
terms of that simple variable, the micro-insurance in Bangladesh is effective. A value of 0 on
the other hand means that it is ineffective
For the aspect of claims of the policyholders, the simple variables and their corresponding values
as per the conclusions are listed in the table below:
Simple variable Conclusion Value
Background Checks on
Claims
Insurance Claim Respondents have claimed
against their insurance. 1
Thorough Background Checks Respondents
believe that their claims have been followed by
proper background checks.
1
Acceptance Rate of Claims More than zero respondents have said that
acceptance rate of claims is high
1
-
24
Denial of Claims Not enough information to make a valid value of
effectiveness
0.5
Time of Claims Not enough information to make a valid value of
effectiveness
0.5
Satisfaction relating to
Claims
Satisfaction with B4 and B6 1
Satisfaction with Amount Realized from Claim 0
Average value 0.7
4.3 FINDINGS REGARDING COLLECTION AND RECOVERY OF PREMIUM
4.3.1 Findings regarding medium of premium collection
From the research, it has been found out that the micro-insurance policyholders usually collect
the premiums through three major ways:
a) Through their offices: The policyholders have to come over to the offices of the
respective micro-insurance company that they are the clients of and have to pay the
premium over the counter to an employee of the company. This is mostly applicable for
urban or semi-urban areas, where setting up an office appears to be useful for the
company.
b) Through agents: For mostly the rural population, the policyholders use this method of
premium collection. Independent agents holding the authorized license provided to them
by the Insurance Development and Regulatory Authority (IDRA) are appointed to go to
these particular locations on selected days and collect the premiums from the
policyholders.
c) Bank accounts: Instead of directly providing the premium to the company, on many
instances, the policyholders pay the amount through a bank. This is advantageous for the
company because this method increases their geographical reach.
A table showing the frequency of the different medium is given below.
Medium Frequency Percentage
Through office 15 37.5%
-
25
Through agents 16 40%
Through bank accounts 6 15%
All of the three mentioned above 1 2.5%
Other methods 2 5%
Conclusion: There are enough methods of premium collection to tackle the different
geographical locations of Bangladesh.
4.3.2 Findings regarding considerations during premium collection
In this section, two different considerations from the perspective of the policyholders while
paying the premiums were reflected upon. Hypotheses tests have been done on both of the
considerations. They are:
a) Consideration of the policyholders occupation (whether the policymakers consider the
occupation of the policyholder, during the process of premium collection)
b) Consideration of the policyholders expense patterns (whether the policymakers consider
the expense patterns of the policyholder, during the process of premium collection)
For these two variables, a bivariate correlation analysis at 1% significance level has been carried
out [Appendix 7] which shows that the correlation coefficient for these two variables is 0.884,
meaning that the correlation between these two variables is quite high.
For consideration regarding the occupation of the policyholders, a hypothesis test was carried out
(the details are in [Appendix 8] under the subheading of Hypothesis 1) where the claim that
the process of premium collection takes into consideration the occupation of the policyholders
was tested.
For that test, the findings are shown in the table below:
Sample mean ( ) 0.40
Significance level for 2 tailed test 6.6%
From the table above, we can see that the significance level for two tailed test is 6.6%. Thus, the
significance level for one tailed test will be 3.3%.
Thus, for any significance level higher than 3.3%, the null hypothesis can be rejected and the
claim that the process of premium collection takes into consideration the occupation of the
policyholders can be established.
-
26
For any significance level lower than 3.3%, the null hypothesis cannot be rejected and the claim
that the process of premium collection takes into consideration the occupation of the
policyholders cannot be established.
For consideration regarding the expense pattern of the policyholders, a hypothesis test was
carried out (the details are in [Appendix 8] under the subheading of Hypothesis 2) where the
claim that the process of premium collection takes into consideration the expense patterns of the
policyholders was tested.
For that test, the findings are shown in the table below:
Sample mean ( ) 0.35
Significance level for 2 tailed test 9.9%
From the table above, we can see that the significance level for two tailed test is 9.9%. Thus, the
significance level for one tailed test will be 4.95%.
Thus, for any significance level higher than 4.95%, the null hypothesis can be rejected and the
claim that the process of premium collection takes into consideration the expense patterns of the
policyholders can be accepted.
For any significance level lower than 4.95%, the null hypothesis cannot be rejected and the claim
that the process of premium collection takes into consideration the expense patterns of the
policyholders cannot be accepted.
Conclusions: At a level of significance of 5%,
It can be concluded that the process of premium collection takes into consideration that
the occupation of the policyholders.
It can be concluded that the process of premium collection takes into consideration that
the expense patterns of the policyholders.
4.3.3 Findings regarding incidence of frauds
The hypothesis tested in this section of the report deals with the incidence of fraud. The claim for
the hypothesis being tested in this scenario is the proportion of population who has been the
victims of fraudulent activities is not 0.
The set of hypothesis for the testing of this claim is given in [Appendix 8] under the subheading
of Hypothesis 3. The frequency distribution of the responses for this question along with the
significance level received through the hypothesis test is given in the table below.
-
27
Has faced fraud in
micro-insurance
Has not faced fraud in
micro-insurance
Frequency 7 33
Proportion p = 0.175 q = 0.825
Significance level for 2 tailed test 0.6%
From the table above, we can see that the significance level for two tailed test is 0.6%. Thus, for
a significance level greater than 0.6%, the null hypothesis can be rejected and the alternate
hypothesis showing the claim that the proportion of population who has been the victims of
fraudulent activities is not 0 can be accepted.
For a significance level lower than 0.6%, the null hypothesis cannot be rejected and we cannot
conclude that the claim is true.
Regarding the awareness of fraud, the following frequency table shows which types of the three
types of fraud: fraud agents, fraud companies and fraud paperwork are the respondents aware of.
Frequency Proportion
Fraud agents 14 0.32
Fraud companies 16 0.36
Fraud paperwork 2 0.05
Missing values 12 0.27
Conclusions:
At a significance level of 5%, it can be concluded that the proportion of population who
has been the victims of fraud is not zero, and is thus significantly large.
27% of the respondents are not aware of any sort of fraudulent activities, which is quite
alarming.
4.3.4 Findings regarding handling of frauds
The respondents seemed to have a generic negative perception of the companies when it came to
how effectively the frauds were being handled. The claim that according to the policyholders,
the companies do not deal with frauds effectively has been tested in this section. The detailed
table generated by SPSS has been given in [Appendix 8] under the subheading of Hypothesis
4.
-
28
A simpler version of that table with the most important points is shown below:
Sample mean ( ) -0.16
Significance level for 2 tailed test 59%
As the significance level for two tailed test is 59%, that for a one tailed test will be 29.5%. Thus,
for significance level of greater than 29.5%, the null hypothesis will be rejected and the claim
that the policyholders believe that the companies do not deal with frauds in an effective manner
will be established.
For significance level of less than 29.5%, the null hypothesis cannot be rejected and the claim
that the policyholders believe that the companies do not deal with frauds in an effective manner
cannot be established.
Conclusion: At a significance level of 5%, it cannot be concluded that the companies do not deal
with frauds effectively.
4.3.5 Findings regarding time of recovery of premium after maturity
From the managerial interviews carried out, the interviewees confirmed that the time of payment
of all the aggregated amount of the premium usually took two months after all the necessary
paperwork were completed by the policyholders subsequent to the maturity of the insurance (if
applicable).
Thus, the mean of the answer to question C8 [Appendix 1] has been tested against the value of
2. The claim being made in this case is that the actual time of recovery of premium is greater
than 2 months after the maturity.
The tables generated by SPSS for the test of this hypothesis and the set of hypotheses being
considered to test the claim made above have been included in [Appendix 8] under the
subheading of Hypothesis 5. A simpler table of the most important values is given below:
Sample mean ( ) 4.11 months
Significance level for 2 tailed test 0.6%
From the table above, we see that the mean from the respondents is 4.11 months. When tested
against the claim being made, the significance level for two tailed test is 0.6%, meaning that the
significance level for one tailed test is 0.3%.
So at a significance level of greater than 0.3%, the null hypothesis can be rejected and the claim
that the actual time of recovery of the premium following the maturity is more than 2 months
-
29
can be accepted. At a significance level of less than 0.3%, the null hypothesis cannot be rejected
and it cannot be concluded that the actual time of recovery of the premium following the
maturity is more than 2 months.
Conclusion: At a significance level of 5%, the claim that the actual time of recovery of
premium following the maturity is greater than 2 months is accepted. The micro-insurance of
Bangladesh is ineffective in context of this simple variable.
4.3.6 Findings regarding satisfaction about amount of premium recovered
The policyholders were asked to mention how satisfied they were on the amount of premium
recovered following the maturity from past experiences, if applicable. With the aggregate result,
a claim that the policyholders are satisfied regarding the amount of premium recovered after
maturity was tested.
The details of the test are present in [Appendix 8] of the report under the subheading of
Hypothesis 6. A table showing the findings of that test has been given below.
Sample mean ( ) 0.16
Significance level for 2 tailed test 55.7%
Thus, we can see from the above table that the significance level for two tailed test is 55.7%.
This means that the significance level for one tailed test is 27.9%.
So, at a significance level of greater than 27.9%, the null hypothesis can be rejected and the
claim that the policyholders are satisfied regarding the amount of premium recovered after
maturity can be accepted.
At a significance level of lower than 27.9%, the null hypothesis cannot be rejected and it cannot
be concluded that the policyholders are satisfied regarding the amount of premium recovered
after maturity.
Conclusion: At a significance level of 5 percent, the claim that the policyholders are satisfied
with the amount of premium that they are recovering after the maturity cannot be accepted.
Thus, in this case, the micro-insurance is ineffective.
4.3.7 Synthesis of micro-insurance premiums
To combine the results from all the simple variables into the parameters, the conclusions from
the individual simple variables have been rated either 1 or 0. A value of 1 means that in
terms of that simple variable, the micro-insurance in Bangladesh.
For the parameter of micro-insurance premiums, the values are given in the table below:
-
30
Simple variable Conclusion Value
Medium of premium
collection
There are three major ways of premium collection,
which is effective in context of Bangladesh.
1
Considerations regarding
premium collection
The process of premium collection does not take into
consideration the occupation of the policyholders.
0
The process of premium collection does not take into
consideration the expense patterns of the policyholders.
0
Incidence of fraud The proportion of population who has encountered fraud
is significantly large.
0
A large proportion of the population is not aware of the
types of fraud that are most common.
0
Handling of frauds It cannot be concluded that the companies do not handle
fraud effectively.
1
Time of recovery of
premiums after maturity
The time of recovery of premium after maturity is
significantly larger than 2 months (which was the
average time claimed by the companies).
0
Satisfaction about amount
of premium recovered
The policyholders are not satisfied regarding the amount
of premium recovered following the maturity
0
Average value 0.25
4.4 FINDINGS REGARDING EMPLOYEES OF MICRO-INSURANCE
4.4.1 Findings regarding instances of misbehavior by employees
In this section we firstly look into whether the policyholders of micro insurance have ever faced
moments where the employees of the company misbehaved. Here we assume that the instances
of misbehavior of employees are zero or less than zero.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 1. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
-
31
Mean Significance Level
Misbehavior by Employees 0.10 0.044
As we can see from the table, the sample mean obtained from the questionnaires for this simple
variable is 0.10. This means that only 10% of the respondents have experienced instances of
misbehavior by employees/agents of the insurance firm.
The level of significance for a two tailed test in this context is 0.044. If we consider a one tailed
test then the significance level would be divided by 2 and result in being 0.022 or 2.2%.
This means that for any level of significance greater than 2.2%, the null hypothesis the mean of
instances of misbehavior by employees/agents is zero will be rejected, while establishing the
alternate hypothesis mean of instances of misbehavior of employees is greater than zero.
Conclusion: With a significance level of 5%, we can conclude that the respondents have
experienced instances of misbehavior by employees or agents. Value of Effectiveness: 0
4.4.2 Findings regarding perception of employee behavior
In this part of the findings we test as to what the perception of policyholders is regarding the
behavior of employees of the insurance firms. For this case we assume that the level of
satisfaction is equal to or below zero signifying that it is unsatisfactory.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 2. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
Mean Significance Level
Perception of Employee
Behavior
1.08 0%
As we can observe from the table that the mean value for the variable is 1.08, signifying that the
respondents have a satisfactory perception towards the behavior of employees. Thus the claim is
that the perception of the policyholders regarding employee behavior is positive.
The level of significance for a two tailed test in this context is 0%. If we consider a one tailed
test, the significance level would be divided by 2 and result in being 0% again brining no change.
-
32
We can conclude here by saying that for any level of significance above 0%, the null hypothesis
the policyholders are not satisfied by the employee behavior and do not have a positive
perception, and the claim that they are indeed satisfied is established.
Conclusion: With a level of significance of 5%, we can conclude that the perception of
policyholders towards employee behavior is satisfactory.
4.4.3 Findings regarding the amount charged by agents
With this section, we test to see whether the amount being charged as commission or premium
rate is acceptable by the respondents or not. For this case we hold the industry standard of agent
commission of 15% and check whether the respondents believe the agents deserve the amount.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 3. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
Mean Significance Level
Perception of Acceptable
Commission Fees
11.50 0.209
From the table we can extract the sample mean for the variable to be 11.50. This means that
according to our respondents the average rate of commission or premium rate should be around
11.5%.
The significance level for a two tailed test in this context is 0.209. If we consider a one tailed test
for the same variable the significance value would be divided by 2 and 0.104 can be obtained.
Thus we can come to the conclusion that for a level of significance greater than 10.4%, the null
hypothesis that the deserved rate of commission for agents is 15% is rejected and the alternate
hypothesis the deserved rate of commission for agents is not 15% stands.
Conclusion: With a significance level of 5%, the null hypothesis is not rejected and we come to
the conclusion that the deserved rate of commission for agents is 15%. For such that any
significance level under 10.4% can be considered the same, so at 5% the commission of 15%
(industry standard) and the sample mean value of 11.5% can be considered the same. Value of
Effectiveness: 1
4.4.4 Findings regarding unrecorded payments
In the insurance industry, unrecorded payments can be held as the precursor for attempts at
fraudulent activities. In this section we test as to whether the respondents have ever faced
-
33
instances of unrecorded payment. Our assumption lies on the hypothesis that the mean of
instances of unrecorded payment is less than or equal to zero.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 4. A simpler table of showing the frequency
distribution of the answers, the mean value and the significance value obtained is given below.
Mean Significance Level
Instances of Unrecorded
Payment
0.18 0.6%
Extracting the sample mean value of 0.18 from the table above we can say the about 18% of the
respondents have faced instances of unrecorded payment. Thus, the claim that we want to test is
that the number of times there were unrecorded payments is significantly greater than zero.
The significance level for the simple variable, in this context, 0.6%. If we convert this two tailed
test into a one tailed test, the level of significance is divided by 2 and we obtain 0.3%. In the end
this means that for any level of significance higher than 0.3%, the null hypothesis is rejected and
the claim is established.
Conclusion: With a significance level of 5%, the null hypothesis is rejected and we can say that
there are instances of unrecorded payments within the policyholders. This means that there are
chances that these respondents are on the stepping stone of an attempt of fraudulent activity.
Value of Effectiveness: 0
4.4.5 Findings regarding knowledge of employees
This section of the findings understands the findings regarding knowledge of the employees. The
test of knowledge has been divided into two parts. The first part understands the perception of
the policyholders towards the knowledge of the micro insurance employees. The second part
deals with the knowledge in a subtle way, by asking as testing whether there have been instances
of failure to answer.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 5 & 6. A simpler table of showing the mean
value and the significance value obtained is given below.
-
34
(a) Employee knowledge
Mean Significance Level
Knowledge of Employees 1.16 0%
The sample mean for the variable as on the table is 1.16. This means that from the sample,
respondents believe that the employees are knowledgeable. Thus, that is the claim that we want
to prove via the hypothesis that we are testing.
The level of significance of a two tailed test in this context is 0%. If we assume a one tailed test
for the variable then the significance level would still be 0%.
We can come to the conclusion that that for any level of significance greater than 0%, we can
reject the null hypothesis the employees are not knowledgeable and establish the claim that
they are in fact knowledgeable.
Conclusion: With a significance level of 5%, we may conclude that the employees in the micro
insurance industry are knowledgeable by rejecting the null hypothesis. The more knowledgeable
the employees are, the more effective the micro insurance scheme. Value of Effectiveness: 1
(b) Failure to Answer
The perception of the policyholders can be biased and gullible, for which we did a second test
which in a subtle manner looks into the fact of knowledge of employees of the micro insurance
companies. In this case we assume that the mean of the failure to answer is equal to zero.
Mean Significance Level
Failure to Answer 0.25 0.1%
The sample mean for the variable as shown above is 0.25. This means that 25% of the
respondents had instances where the employee was unable to answer the queries of the
policyholders.
The significance level for a two tailed test of the variable is 0.001 in this context. If we consider
a one tailed test the level of significance will become 0.05% for the variable.
We can now come to a conclusion that any level of significance above 0.05% will cause the null
hypothesis mean of failure to answer is equal to zero to be rejected and the alternate hypothesis
mean of failure to answer is greater than zero to be established.
-
35
Conclusion: With a significance level of 5%, we can say the null hypothesis is rejected, meaning
that there are instances of failure to answer by the employees of the micro insurance companies.
This shows as to how, the perception of the policyholders may have been wrong. This renders the
industry as ineffective.
4.4.6 Findings regarding frequency of visits made by agents
In this section we test whether the frequency of visit made by agents is high or low for those
policyholders who have to pay the premiums via agents. Here we assume that the agents do not
visit the policyholders regularly.
The set of hypothesis taken into consideration and detailed tables generated from SPSS are given
in [Appendix 9], under the subtopic Hypothesis 7. A simpler table of showing the mean value
and the significance value obtained is given below.
Mean Significance Level
Regular Visits of Agents 0.87 0%
The sample mean for the variable as shown on the table is 0.87. This means 87% of the
respondents have agents who regularly visit the policyholders. Thus, the claim that we want to
test is that a significantly large proportion of policyholders believe that agents visit regularly.
The significance level for a two tailed test in this context is 0%. If we consider a one tailed test
for the variable, the level of significance would still be 0%.
Thus we can conclude that for at any level of significance above 0%, we reject the null
hypothesis and establish the alternate hypothesis that a significantly large proportion of
policyholders believe that agents make regular visits.
Conclusion: With a level of significance of 5%, rejecting the null hypothesis, we can claim that
the agents regularly visit the policyholders making the micro insurance scheme effective in
Bangladesh in this context.
4.4.7 Synthesis of micro-insurance employees
Simple variable Conclusion Value
Instances of Misbehavior by
Employees/Agents
The policyholders have experienced instances
where agents/employees misbehaved with them.
Rendering the insurance ineffective.
0
-
36
Perception of Employee
Behavior
The policyholders believe that the employees
behavior is satisfactory 1
Amount of commission
charged by Agents
The mean commission suggested by policyholders
is the same as the 15% being charged currently in
the industry causing this section to be effective.
1
Instances of Unrecorded
Payments
Policyholders have faces instances where the firm
or its agents have not recorded the payments. This
renders this section as ineffective.
0
Knowledge of Employees
Policyholders believe that knowledge of
employees regarding micro insurance schemes is
satisfactory.
1
However there have been several instances of
failure to answer queries by the agents which
causes this section to be ineffective.
0
Frequency of Visits made by
Agents
87% of the policyholders have agents who visit
regularly to collect the premiums. 1
Average value 0.60
4.5 FINDINGS REGARDING COVERAGE OF MICRO-INSURANCE
In this part of the report, we try to find out as to whether micro insurance is able to cover most of
the variety of insurance needs of the people of Bangladesh.
The table below shows how the different items need to be insured, are already insured, or both,
in the perspective of our sample respondents.
Items
Assets that need
to be insured
(Frequency)
Assets that are
insured
(Frequency)
Both
(Frequency)
Value of
Effectiveness
1) Hospitalization 5 3 1 0
2) Primary health care 5 - 2 0
3) Maternity 3 - - 0
-
37
4) Life Insurance 4 15 19 1
5) Retirement savings plans 5 1 2 0
6) Against weather 4 1 - 0
7) Against rising cost of cultivation
- - - 0
8) Against pest attacks and diseases
1 - - 0
9) Livestock - - - 0
10) Housing 3 1 - 0
11) Vehicles 7 1 - 0
Total 37 22 24 1
Average Value of Effecti