December 11, 2011
Dr. Cesar Rufino - ECONMET Page
“ANALYSIS ON THE FACTORS AFFECTING POPULATIONS BELOW POVERTY LINE”
An Empirical Paper
Presented to
The Faculty of the School of Economics
De La Salle University
In Partial Fulfillment of the Requirements for the Course
ECONMET
Submitted by:
Co, Mark Anthony F.
Submitted to:
Dr. Cesar Rufino
December 11, 2011
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Table of Contents:
Part I:
I. Introduction
a) Background of the Study
b) Statement of the Problem
c) Objective of the Study
d) Significance of the Study
e) Scope, Limitations and Nature of the Study
II. Review of Related Literature
Part II:
III. Operational Framework
a) Description of the Variable Used
b) Hypothesized Relationship of Variables
c) Introduction of Hypothesized Economic Model
IV. Methodology
a) Data
b) Estimation and Inference Procedures
6
9
9
10
11
13
22
24
26
28
28
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
V. Empirical Results and Interpretation
a) Estimated Values of OLS Regression
b) Test for Multicollinearity
c) Test for Heteroscedasticity
d) Test for Misspecification (Specification of Bias)
e) Corrected Model and OLS Estimates
f) Test for Significance of the OLS Estimates
g) The Final Model and OLS Final Estimates
VI. Conclusion
VII. Bibliography
VIII. Appendix
30
36
40
43
43
44
46
48
50
52
Acknowledgement:
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
I, the researcher of this project would like to extend my deepest and sincerest gratitude
to Dr. Cesar Rufino for his unwavering support and for giving sufficient knowledge and
information on how to work on to the project effectively and equipping me the necessary skills
and techniques for this paper; also for tackling the different processes and imparting their
knowledge and time to fully understand the project. This empirical paper and research is not
that plain sailing to do. Assistance and guidance of others are needed to make this empirical
paper possible. I, the researcher would also like to thank each and every member of the class
and contributors for their support and cooperation to make this project a meaningful one. To
my mentors and friends who shared their time to guide me in this empirical paper project and
for giving their moral as well as emotional support. And lastly to God for the blessings and
inspiring each and every members of the class to make their successful paper possible, for
bestowing his guidance and love to be the inspiration in everything.
Every member of this class in ECONMET is capable of handling and doing the project. I,
the researcher would like to thank the people for cooperating and maintaining the camaraderie
for the success of this project. May your sacrifice and helping hands continue to be glorified and
bless by God.
Abstract:
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
This empirical research is all about the effect of crime rate, adult literacy rate, GDP per
capita, corruption index, unemployment, and if a country is develop or not on population living
below the poverty line among countries around the world both developed and developing
countries. From the model itself, which has been done and on the different test that was done
2 out of the 6 exogenous variables are significant which is the crime rate and the
unemployment rate. The other 4 variables of literacy rate, GDP per capita, corruption index and
the dummy variable if a country is developed are deemed to be insignificant to the model.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Background of the Study
More than one billion people in the world live on less than one dollar or P43 Philippine
Peso a day, another 2.7 billion people struggle to survive on less than two dollars or P86
Philippine Peso per day (Smith, 2009). Poverty in the developing world, however, goes far
beyond income poverty. It means having to walk more than one mile everyday simply to collect
water and firewood; it means suffering diseases that were eradicated from rich countries
decades ago. Every year eleven million children die most under the age of five and more than
six million from completely preventable causes like malaria, diarrhea and pneumonia.
In some deeply impoverished nations less than half of the children are in primary school
and fewer than 20 percent go to secondary school. Around the world, a total of 114 million
children do not get even a basic education and 584 million women are illiterate (Williams,
2007). Following are basic facts outlining the roots and manifestations of the poverty affecting
more than one third of our world:
Health
Every year six million children die from malnutrition before their fifth birthday.
More than 50 percent of Africans suffer from water-related diseases such as cholera
and infant diarrhea.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Everyday HIV/AIDS kills 6,000 people and another 8,200 people are infected with this
deadly virus.
Every 30 seconds an African child dies of malaria-more than one million child deaths a
year.
Each year, approximately 300 to 500 million people are infected with malaria.
Approximately three million people die as a result.
TB is the leading AIDS-related killer and in some parts of Africa, 75 percent of people
with HIV also have TB.
Hunger
More than 800 million people go to bed hungry every day, 300 million are children.
Of these 300 million children, only eight percent are victims of famine or other
emergency situations. More than 90 percent are suffering long-term malnourishment
and micronutrient deficiency.
Every 3.6 seconds another person dies of starvation and the large majorities are
children under the age of 5.
Water
More than 2.6 billion people-over 40 per cent of the world's population-do not have
basic sanitation, and more than one billion people still use unsafe sources of drinking
water.
Four out of every ten people in the world don't have access even to a simple latrine.
Five million people, mostly children, die each year from water-borne diseases.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Agriculture
In 1960, Africa was a net exporter of food; today the continent imports one-third of its
grain.
More than 40 percent of Africans do not even have the ability to obtain sufficient food
on a day-today basis.
Declining soil fertility, land degradation, and the AIDS pandemic have led to a 23
percent decrease in food production per capita in the last 25 years even though
population has increased dramatically.
For the African farmer, conventional fertilizers cost two to six times more than the
world market price.
Poverty in the Philippines
The Population Commission said there are 30.6 million Filipinos or 6.12 million families
who are suffering from poverty. When I learned about this, I took consolation with the notion
that I am more fortunate and should not waste the things that I have, yet I felt dismayed over
the complacency of our national government officials who seem undisturbed by the fact that 40
percent of their constituents live below the poverty line throughout the country's 78 provinces,
84 cities or 41,940 barangays.
The poorest of the poor are the indigenous peoples, small-scale farmers who cultivate
land received through agrarian reform, landless workers, fishers, people in upland areas and
women.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
a) Statement of the Problem
Poverty is everywhere, it is present in all countries even to those we say that are
industrialized already, it is has been a world-wide discussion on how we can combat this
problem that has been here for almost the start of history. Many people try to do their best
to get out of these status of not being able to sustain their daily needs and of course their
wants in life, it’s been an endless topic in most political events and campaigns, that they
would end this problem. So I would like to know now what certain factors affect the poverty
line that contributes to one of the greatest phenomenon of the world.
This paper now would like to answer the question of whether there exists a relationship
between the poverty line and the following explanatory variables: crime rate, literacy rate,
GDP per capita, corruption in the government, unemployment rate and a country that is
already developed and those that are still developing. This study now would like to know if
these key variables truly have an effect on the world-wide problem of poverty.
b) Objectives of the Study
This paper has the following objectives:
i. To construct an appropriate econometric model that will be relevant and applicable for
the poverty line trend in the world-wide region.
ii. To determine whether crime rate, literacy rate, GDP per capita, corruption in the
government, unemployment rate and a country that is already developed and those
that are still developing are significant indicators of poverty line in the world-wide
region by using econometric theories and principles.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
iii. To give an insight on what the countries in this region can possibly do in order to battle,
combat and lower down the poverty line of their respective country.
iv. To determine if there are key aspects that would later result to a more devastating
increase in the poverty line that each country or the world prevents.
v. To understand our own position in the poverty line on how we, La Sallians should be
grateful that we are living in a comfortable life and is not that much affected by these
world-wide phenomenon.
c) Significance of the Study
This study would be very useful for those countries that want to know if these key
factors could really influence the poverty line of their own respective country, to the
government establishments, especially those which are concerned with economic development
and improvement of the status of their people. The people heading these departments can take
their cue from this research on what can decrease the population below the poverty line, or in
other words, the standard of living. Through this research, countries world-wide could focus
their policies towards those variables which are vital in decreasing the population under the
poverty line that would later on results to a more stable and developed country that would, not
waste valuable resources on improving those factors which either way cannot increase the
country’s welfare. They could prioritize on these key concepts and factors that would really
solve the different problems of their nation.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
This study would also be useful for those people who are ignorant of other people’s
sufferings and current status that they could not even sustain their own daily needs for them to
be able to survive their daily routine in life, how much more to satisfy their wants. Some of the
people out there are clueless to what is really happening in the global outlook especially in
Africa and some parts of Asia. This study would like to encourage these people to be satisfied
and love what they already have and by not wishing and yearning for more to what they
already have.
d) Scope, Limitations and Nature of the Study
The focus of the study is to find out if crime rate, literacy rate, GDP per capita,
corruption in the government, unemployment rate and a country that is already developed and
those that are still developing are the key factors that population below poverty line in
countries. Some people believe that the crime rate has a positive effect on poverty rate, as a
country’s crime rate increases so is the population living below the poverty line, next is the
literacy rate, the more people that has the basic knowledge to read and write affects the crime
rate because if these people are knowledgeable on this matter they would try to do their best
to find the job that would suit them and would prevent themselves from committing these
crimes. Next is the GDP per capita, as a countries GDP per capita increase it decreases the
poverty line because more products and services are made over population thus affecting the
overall poverty rate of a country. Next in the corruption index, it is said that if the government
officials are corrupt and are stealing the money of the people through their port barrels and
supposed budgets that should have been spent for the benefit of the people. If these problems
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
are solved upon, poverty rate would decrease because the money now that has not been stolen
by these officials would result a greater and more prosperous kind of living for its people. Last is
if a country is developed or not, developed countries tends to make more blue-collar crimes
and in the developing countries, physical and theft are the basic crimes that are predominant.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
1. Crime Rate
A study by McClatchy Newspapers (2007) released in March finds that the ranks of the
severely impoverished are rapidly escalating. The study found that the percentage of poor
people who are living in extreme poverty has reached a 32 year high. Today nearly 2 billion
people live in deep or severe poverty. This is defined as individuals living at half of the poverty
line. This drastic rise in the level of poverty extends beyond the traditional ghetto and reaches
to suburban and rural communities.
The relationship between poverty and crime has been a controversial subject over the
years. Many scholars argue that poverty does not have a causal relationship to crime because
there are countries in which poverty is very high but the crime rate is relatively low. It would
say that in this country it would be hard to argue that there is not a relationship between crime
and poverty. Poor people make up the overwhelming majority of those behind bars as 53% of
those in prison earned less money per year before incarceration.
Sociologist and criminal justice scholars have found a direct correlation between poverty
and crime. One economic theory of crime assumes that people weigh the consequences of
committing crime (Williams, 2008). They resort to crime only if the cost or consequences are
outweighed by the potential benefits to be gained. The logical conclusion to this theory is that
people living in poverty are far more likely to commit property crimes such as burglary, larceny,
or theft.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
The rising levels of poverty, then, should alarm those of us engaged in ministry to
prisoners, ex-prisoners and their families. It follows that as goes the poverty rates, so go the
crime rates and subsequently the prison rates. If the relationship between poverty rates and
crime rates holds, and suspected that it will, we can expect to be faced with the challenge of
ministering to even higher numbers of inmates and former inmates. Those of us who minister
to men and women in transition from prison and the families of inmates can expect to have our
meager resources taxed to the limit.
Establishing satisfying employment and economic well-being are important factors for
successful reintegration from prison to the community. We who are engaged in this ministry
are being forced to be more innovative than we ever have been in order to effectively minister
to former prisoners and their families. We must identify considerably more resources than
ever. We must reach out to a wider network of supporters than ever before to make our case
for support.
Ironically, as the numbers of those in extreme poverty has increased so has the number
of those who have become wealthy. Bridges need to be formed between those who minister in
cities and other impoverished areas with meager resources and those who possess significant
financial resources. Those who are interested in helping must adopt the attitude of teaching
people how to fish instead of passing out fish sandwiches if persistent problems such as crime
and poverty are to be effectively addressed.
All of society benefits when the least of these are helped to establish or regain dignity
by elevation from poverty and crime to lives characterized by work and productivity. It will take
all of us working together to make a real impact on this daunting problem.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
2. Literacy Rate
The relationship between level of education and poverty is clear. The National Institute
for Literacy estimates that 43% of adult with very low literacy skills live in poverty. About 70%
of adult welfare recipients have lower level literacy skill on the National Assessment of the
Adult Literacy. About 47% of adult welfare recipients have not graduated from high school.
Individuals ages 25-34 who dropped out of high school are more than three times as likely to
receive public assistance as high school graduates who did not go on to college. According to
one study of welfare recipients without high school diplomas, when recipients increase their
basic skills, they tend to make substantial improvements in employment, earnings, and self-
sufficiency. In a study of mothers receiving welfare, each additional year of schooling led to
approximately a 7% wage increase.
People from poor families as well as the long-term unemployed, seniors, native people,
prisoners, people with disabilities, and racial and cultural minorities all have lower rates of
literacy and higher rates of poverty. They have fewer choices in jobs, education, housing and
other things we need to have full lives. Poverty and low literacy form a cycle that is difficult to
break.
According to a 2003 Report Card on Child Poverty Campaign 2000, even in the year
2001, almost one in six children in developed countries still lived in poverty. These children are
often not well served by the school system where they are likely to be labeled and placed in
classes where less is expected of them and less may be offered. Their parents may not have the
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
information, confidence, or skills to challenge the school system for help. Many poor children
either drop out of high school or graduate without being fully literate.
3. GDP per Capita
Finding a suitable measure which captures the depth and breadth of poverty, at a
national and individual level is an enormous challenge to researchers. There are difficulties with
each of the following commonly used methods. Collecting data is complicated and costly, so
accuracy and comparisons between countries may not be reliable.
The most common way to measure a country's wealth or poverty is Gross Domestic
Product (GDP) or Gross National Income (GNI) per capita. GDP measures the value of the goods
and services that a country produces. GNI is slightly different, because it measures the value of
goods and services produced by the assets that a country and its citizens own, even if the
production takes place elsewhere. So GDP/GNI tells us how much income a country has in total,
but it doesn't work perfectly as a measure of poverty and wealth. This is because it measures
costs which may not impact on a household level, and it fails to consider such costs as unpaid
labor and environmental damage.
A country's total GDP or GNI is often divided by the population to give an indication of
the average income per person; however, this gives no indication of actual distribution of the
income within a nation. A few people may be extremely rich while the majority may be very
poor.
Also, international comparisons require conversions into a common currency (usually
US$) using the average official exchange rate reported by the International Monetary Fund. This
may produce inaccurate results, and does not take proper account of how prices vary between
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
countries. Sometimes GDP or GNI figures are adjusted to take account of price differences; this
is known as PPP (purchasing price parity). It is very difficult, however, to measure accurately the
differences between countries or communities where consumption patterns are very different.
The figure also does not measure how well-off people are in terms of their human
development or standard of living. Subsistence farmers may be able to provide most of their
needs even though they contribute only a very small amount to the GDP/GNI.
4. Government Efficiency or Corruption Index
The issue of corruption resonates in developing countries. In the Philippines, for
instance, the slogan of the coalition that is likely to win the 2010 presidential elections is
"Without corrupt officials, there are no poor people.
Not surprisingly, the international financial institutions have weighed in. The World Bank
has made good governance major thrust of its work, asserting that the World Bank Group focus
on governance and anticorruption follows from its mandate to reduce poverty; a capable and
accountable state creates opportunities for poor people, provides better services, and improves
development outcomes.
Corruption often conjures up images of people getting rich. But in fact, corruption's
connections to poverty are far more numerous and pervasive. Corruption delays, distorts and
diverts economic growth. It comes in a variety of forms, and while no two countries are alike,
there are common dilemmas for all to see.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
The links between corruption and poverty affect both individuals and businesses, and
they run in both directions: poverty invites corruption, while corruption deepens poverty.
Corruption both causes and thrives upon weaknesses in key economic, political and social
institutions. It is a form of self-serving influence akin to a heavily regressive tax, benefiting the
haves at the expense of the have-nots. Trust essential to financial markets and effective
governments everywhere is difficult to build in poor and corrupt societies.
Poor people and economically strapped businesses have few economic alternatives, and
where serious corruption is the norm, they are even more vulnerable to exploitation. In that
sense, there is no such thing as petty corruption: police shakedowns in a public market, or
roadblocks in the countryside where farmers must pay up in order to transport produce to the
city, may yield seemingly trivial sums of money, but they help keep poor people poor.
Low-level officials themselves may have trouble earning an honest living. In poor
societies, they are often underpaid, when they are paid at all, and must provide a stream of
payments to patrons at higher levels. In such settings, bribery, extortion and theft become
matters of survival.
For businesses small and large, and particularly for international investors, serious
corruption has formidable costs. It is tempting--and, at times has been fashionable--to think of
bribery as grease for the wheels of bureaucracy. And indeed a sweetener or backhander paid to
the right person at the right time might help one firm, on one day, get a permit or a license
more quickly.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
5. Unemployment Rate
Unemployment and poverty are the two major challenges that are facing the world
economy at present. Unemployment leads to financial crisis and reduces the overall purchasing
capacity of a nation. This in turn results in poverty followed by increasing burden of debt. Now,
poverty can be described in several ways. As per the World Bank definition, poverty implies a
financial condition where people are unable to maintain the minimum standard of living.
Poverty can be of different types like absolute poverty and relative poverty. There may
be many other classifications like urban poverty, rural poverty, primary poverty, secondary
poverty and many more. Whatever be the type of poverty, the basic reason has always been
lack of adequate income. Here comes the role of unemployment behind poverty. Lack of
employment opportunities and the consequential income disparity bring about mass poverty in
most of the developing and under developed economies of the world.
It is true that unemployment and poverty are mostly common in the less developed
economies. However, due to the global economic recessions, the developed economies are also
facing these challenges in the recent times. The US subprime crisis and its wide spread impacts
have played a major role in worsening the situation.
In India, the problems of unemployment and poverty have always been major obstacles
to economic development. Underemployment and unemployment have crippled the Indian
economy from time to time. Even during the period of good harvest, the Indian farmers are not
employed for the entire year. Excessive population is another major problem as far as Indian
economy is concerned. Regional disparity is also crucial in this context. A part of the urban
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
workforce in India is subjected to sub-employment. Mass migration from rural to urban regions
is adding to the problems of unemployment and poverty in India.
Economic reforms, changes in the industrial policy and better utilization of available
resources are expected to reduce the problem of unemployment and poverty that results from
it. The economic reform measures need to have major impacts on the employment generating
potential of the economy. The governmental bodies are also required to initiate long term
measures for poverty alleviation. Generation of employment opportunities and equality in
income distribution are the two key factors that are of utmost importance to deal with the dual
problem of unemployment and poverty.
6. Developed Vs. Developing Countries
What does it mean to be poor? How is poverty measured? Third World countries are
often described as “developing” while the First World, industrialized nations are often
“developed”. What does it mean to describe a nation as “developing”? A lack of material
wealth does not necessarily mean that one is deprived. A strong economy in a developed
nation doesn’t mean much when a significant percentage even a majority of the population is
struggling to survive. Successful development can imply many things, such as: An improvement
in living standards and access to all basic needs such that a person has enough food, water,
shelter, clothing, health, education, etc; A stable political, social and economic environment,
with associated political, social and economic freedoms, such as equitable ownership of land
and property; The ability to make free and informed choices that are not coerced; Be able to
participate in a democratic environment with the ability to have a say in one’s own future;
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
To have the full potential for what the United Nations calls Human Development;
Human development is about much more than the rise or fall of national incomes. It is about
creating an environment in which people can develop their full potential and lead productive,
creative lives in accord with their needs and interests. People are the real wealth of nations.
Development is thus about expanding the choices people have to lead lives that they value. And
it is thus about much more than economic growth, which is only a means if a very important
one of enlarging people’s choices.
Relative poverty measures are used as official poverty rates in several developed
countries. As such these poverty statistics measure inequality rather than material deprivation
or hardship. The measurements are usually based on a person's yearly income and frequently
take no account of total wealth. The main poverty line used in the OECD and the European
Union is based on economic distance, a level of income set at 60% of the median household
income.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
a) Description of the Variables Used
Before proceeding to the study and analysis of the model itself, the components of the
model need first be described. These components are comprised of both the dependent and
independent variables. The independent variables are the so called exogenous variables of the
model and are thus not affected by any of the other variables inside the model unlike the
dependent or endogenous variable. The following table will briefly explain what each variable in
the model stands for:
Table 1.) Variable Names and Descriptions
Variable Definition
Poverty The Poverty variable is a quantitative variable pertaining to the population under poverty line present height in countries world-wide.
Crime The Crime variable is a quantitative variable which pertains to the total crime over
population or crime mean of each world-wide country.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Literacy The Literacy variable is a quantitative variable which pertains to the overall literacy or
capability of a person to read and write of the total population of each world-wide country.
Gdp The Gdp variable is a quantitative variable which pertains to the GDP per capita, the
value of all final goods and services produced within a country in a given year divided by the
average population for the same year.
Corindx The Corindx variable is a quantitative variable which pertains to the corruption index, the
annual ranking of countries by their perceived levels of corruption, as determined by expert
assessments and opinion surveys of each world-wide country. The higher the Corindx the cleaner the country of corruption, the
lower the Corindx, the higher the probability that corruption is prevalent.
Unempl The Corindx variable is a quantitative variable which pertains to the unemployment rate, the measurement of total unemployed workers /
the labor force of each country.
Developed The Developed variable is a dummy variable which pertains to the present industrial status of each country world-wide. It is given a value
of 1 if the country is developed and 0 if otherwise.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
b) Hypothesized Relationship of Variables
Table 2.) A-Priori Expectations of the Regressors on the Regressand
Endogenous Variable/ Regressand: Poverty (present rate of population living under poverty line
ExogenousVariables/Regressors
A-Priori Expectations on the Relationship of the Regressors on the Regressand
Positive / Negative effect of A-Priory
Expectation
Crime It is said that the crime rate of each country is directly related to the poverty
level of each country, the higher the population of the country living under the poverty line is, the higher the probability
that they would depend on crimes to earn money, the higher the crime rate is, the higher or many people are said to be living under poverty line and thus; cannot sustain their daily needs to survive each
day of their life.
+
Literacy It is said in a study made by (Graves, 2007) the literacy rate or the ability of a
person to read and write in their own native language, generally affects the poverty line of country, the higher the
literacy rate of a country is the lower the people living in poverty line and vice versa, the lower the literacy rate, the higher the population living under the
poverty line.
-
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Gdp Generally, the GDP per capita is the PPP (or purchasing price parity) that the value
of all final goods and services produced within a country in a given year divided by the average population for the same
year, this means that the higher the GDP per capita of each country is that people
are producing goods and offering services that corresponds to the total population, meaning those with the higher GDP per capita, it is presumed that people are
working thus earning money, and in the middle-run would affect poverty as a
whole of a specified country.
-
Corindx One factor that is said that really affects the poverty level and the population
living under poverty is the corruption of the government officials in the public sector, and corporate leaders in the private sectors. This affects now the money that should have been given
prioritized to those people who are in need of support and service from the
government, the higher the corruption index of a country, it is the higher the
speculation of people that the government itself is stealing from its own funding that should have been spent to
create or improve one’s country. This has a negative effect on our Regressand
because corruption increases the poverty level of population living under the
poverty line.
-
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Unempl The unemployment rate has a positive direct effect on the poverty line because
if people are not working, or does not have a job, they cannot earn money that could sustain themselves to survive the
living here in the country; thus this has a positive effect on the population level
living under the poverty line.
+
Developed Developed Countries are said to be more advance and industrialized in a way that they could easily sustain themselves to survive and has a lower unemployment
rate and high corruption index.
-
c) Introduction of Hypothesized Economic Model
The econometric model made for this study is based upon the intuition, theory, and
research, studies of researchers, facts, concepts, and claims of previous related works on the
matter. The level of population living under poverty line will now be under the study as to
whether the exogenous variables above are true determinants of the level of population living
under poverty line. The primary focus of this study is finding out which of this Regressors would
have a positive and negative effect on the Regressand, it is also to find out if our A-Priory
expectations are close or true compared to the corrected estimates of each variable.
The model that will be used for this research is a linear-linear model (lin-lin model).
Population living under poverty line will be represented by Poverty; total crimes over
population by ‘crime’; adult literacy rate by Literacy; GDP per capita by Gdp; corruption in the
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
government by Corindx; unemployment in the country by Unempl; and lastly the dummy
variable if a country is already developed by Develop.
The null and alternative hypotheses for this regression analysis are the following:
Total Crime / Population - H0: β1 = 0 vs. H1: β1 ≠ 0
Adult Literacy Rate - H0: β2 = 0 vs. H1: β2 ≠ 0
GDP per capita - H0: β3 = 0 vs. H1: β3 ≠ 0
Corruption in the Government - H0: β4 = 0 vs. H1: β4 ≠ 0
Unemployment Rate - H0: β5 = 0 vs. H1: β5 ≠ 0
Developed - H0: β6 = 0 vs. H1: β6 ≠ 0
The estimated econometric model is illustrated:
The Model on the Determinants of Level of Population under Poverty Line for Estimation:
Poverty = β1 + β2Crime - β3Literacy - β4Gdp - β5Corindx + β6Unempl - β7Develop + u
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
a) Data
The data gathered was based on 75 countries composed of the different continents over
the world, I picked these 75 countries world-wide, not just in a single continent to be able to
determine how poverty line is produced in different segments of society, both the developed
and the developing countries, each data factors was researched hand by hand by the
researcher. The researcher picked the most recent data to assure that the case study would be
recent and applicable to modern day activities and phenomenon.
b) Estimation and Inference Procedures
Multiple variable regression will be employed by this study in order to find out what
truly affects the population level under poverty line. I, the researcher would then use the
selected and proposed software programs specializing in estimation and solving like Stata and
Gretl in approximating the coefficients, R-Squared and to determine which of the independent
variables and dummy variables have significant effect on the Regressand variable. The model
that will be used is of the linear-linear model (lin-lin model). The method of estimation that will
be applied is Ordinary Least Squares and the type of data that would be used is the cross
section data. Upon the regression of the model, it will be put in various tests. Since the data is
cross-sectional, the violations to test for are multicollinearity and heteroscedasticity. We would
also determine the normality of the data and how significant each variable really is to the
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
endogenous variable. To test for multicollinearity, auxiliary regressions and the variance
inflation factor would be used. In testing for heteroscedasticity, both the Breush Pagan Cook
Weisberg Test and White’s Test will be considered. If these problems persist in the data,
remedies and offset of variables should be done to make the analysis good and significant. Test
for misspecification or specification bias will also be executed through the Ramsey Reset Test.
Upon the finishing of the test the final estimates and final model would be deemed finished.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
a) Estimated Values of OLS Regression
_cons 36.01429 13.72835 2.62 0.011 8.619801 63.40879 developed 1.231924 4.263661 0.29 0.774 -7.27608 9.739928unemployment .5870588 .1103334 5.32 0.000 .366892 .8072256corruption~x -1.201231 1.038222 -1.16 0.251 -3.272971 .8705093 gdp -.0000779 .0001119 -0.70 0.489 -.0003013 .0001455 literacy -.1699038 .1460657 -1.16 0.249 -.4613734 .1215657 crime .2572699 .119633 2.15 0.035 .018546 .4959938 poverty Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 18629.7625 74 251.753548 Root MSE = 11.315 Adj R-squared = 0.4915 Residual 8705.56469 68 128.02301 R-squared = 0.5327 Model 9924.19784 6 1654.03297 Prob > F = 0.0000 F( 6, 68) = 12.92 Source SS df MS Number of obs = 75
Model 1: OLS, using observations 1-75Dependent variable: poverty
Coefficient Std. Error t-ratio p-valueconst 36.0143 13.7283 2.6234 0.01074 **crime 0.25727 0.119633 2.1505 0.03507 **literacy -0.169904 0.146066 -1.1632 0.24881gdp -7.7896e-05 0.000111949 -0.6958 0.48891corruptionindex -1.20123 1.03822 -1.1570 0.25132unemployment 0.587059 0.110333 5.3208 <0.00001 ***developed 1.23192 4.26366 0.2889 0.77351
Mean dependent var. 22.09933 S.D. dependent var. 15.86674Sum squared residual 8705.565 S.E. of regression 11.31473R-squared 0.532707 Adjusted R-squared 0.491475F(6, 68) 12.91981 P-value(F) 1.10e-09Log-likelihood -284.7040 Akaike criterion 583.4080Schwarz criterion 599.6304 Hannan-Quinn 589.8854
OLS Regression using the Gretl software
OLS Regression using the Stata 11 software
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
This table was the result after I have done regression with Poverty with the different
quantitative and dummy variables that are considered to be exogenous or Regressors of the
endogenous variable. The R-Squared for this regression is 0.5327, signifying that 53.27% of the
variation in population below poverty line is explained by either any of the exogenous variables.
We could say that the R-Squared is good because it is more than half of the endogenous
variables can be explained by these exogenous variables. The sample regression line fits the
data quite well. Now let’s move on to the co-efficient of each exogenous variable and the
constant so that we could view on how these factors can influence and affect the endogenous
variable “Poverty”.
The coefficient of crime over population is 0.2572699, signifying that as total crime over
population increases by 1, population under poverty line will increase by
approximately .2572699 base points. The computed t is equal to 2.15, which is less than the
critical t. At the 95% confidence level, we could see that P>|t| is 0.035 which is less than 5%,
meaning that the total crime over population is significant to the endogenous variable which is
Poverty. My A-Priory expectation of a positive effect is deemed correct by this OLS regression.
Now let us move on to the next exogenous variable which is literacy rate, the coefficient
of literacy rate is -.1699038, signifying that as the literacy rate of a country’s population
increases by 1, population under poverty line will decrease by approximately .1699038 base
points. The computed t is equal to -1.16, which is less than the critical t. At the 95% confidence
level, we could see that P>|t| is 0.249 which is greater than 5%, meaning that the literacy rate
of a country’s population is not significant to the endogenous variable which is Poverty. My A-
Priory expectation of a negative effect is deemed correct again by this OLS regression.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Now let us move on to the next exogenous variable which is GDP per capita, the
coefficient of GDP per capita is -.0000779, signifying that as the GDP per capita of a country’s
increases by 1, population under poverty line will decrease by approximately .0000779 base
points. The computed t is equal to -0.70, which is less than the critical t. At the 95% confidence
level, we could see that P>|t| is 0.489 which is greater than 5%, meaning that the GDP per
capita of a country is not significant to the endogenous variable which is Poverty. My A-Priory
expectation of a negative effect is deemed correct again by this OLS regression.
Now let us move on to the next exogenous variable which is inefficiency in the
government sectors that could lead to corruption, the coefficient of corruption index is
-1.201231, signifying that as the corruption index of a country’s government increases by 1,
population under poverty line will decrease by approximately 1.201231 base points. The
computed t is equal to -1.16, which is less than the critical t. At the 95% confidence level, we
could see that P>|t| is 0.251 which is greater than 5%, meaning that the corruption index of a
country’s government is not significant to the endogenous variable which is Poverty. My A-
Priory expectation of a negative effect is deemed correct again by this OLS regression.
Now let us move on to the next exogenous variable which is unemployment rate, the
coefficient of unemployment rate is .5870588, signifying that as the unemployment rate of a
country’s population increases by 1, population under poverty line will increase by
approximately .5870588 base points. The computed t is equal to 5.32, which is really greater
than the critical t. At the 95% confidence level, we could see that P>|t| is 0.00001 which is
really less than 5%, meaning that the unemployment rate of a country’s population is very
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
significant to the endogenous variable which is Poverty. My A-Priory expectation of a positive
effect is deemed correct again by this OLS regression.
Now let us move on to the last exogenous variable which is also a dummy variable
which is if a country is a developed country, the coefficient of develop is 1.231934, signifying
that as the development of a country increases by 1, population under poverty line will increase
by approximately 1.231934 base points. The computed t is equal to 0.29, which is less than the
critical t. At the 95% confidence level, we could see that P>|t| is 0.774 which is greater than
5%, meaning that the developed country is not significant to the endogenous variable which is
Poverty. My A-Priory expectation of a negative effect is deemed wronged by this OLS regression
because it has a positive effect to the increase of population under poverty line.
The intercept coefficient is equal to 36.01429. If the exogenous variables of crime,
literacy, gdp, unemployment, corruption index, and if a country is develop are all equal to 0,
the population below poverty line would be at 36.01429 base points. However, since the
computed t is less than the critical t, it is insignificant at the 95% confidence level, also having a
P>|t| of 0.011.
INITIAL MODEL FROM PRIMARY OLS REGRESSION:
Poverty = 36.01429 + .257699Crime - .1699038Literacy - .0000779Gdp – 1.201231Corindx + .5870588Unempl + 1.231924Develop + u
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Tests for Classical Linear Regression Model Violations
The characteristics for the OLS regression for a Classical Linear Regression Model that
were earlier on simply assumed are now to be relaxed. These include the normality and
homoscedasticity of the stochastic random variable, the absence of perfect multicollinearity,
and the absence of autocorrelation between errors, wherein we would test on later.
Test for the Normality of the Stochastic Random Variable
0
0.01
0.02
0.03
0.04
0.05
0.06
-30 -20 -10 0 10 20 30 40 50
De
nsi
ty
uhat2
uhat2N(3.9198e-015,11.315)
Test statistic for normality:
Chi-square(2) = 22.547 [0.0000]
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Testing for the normality of the stochastic random variable simply means checking
whether the error terms per observation actually follows a normal distribution as a whole. The
null hypothesis is that the error terms or the random stochastic random variables per
observation follow the normality distribution assumption, and the alternative hypothesis is that
the normality distribution assumption does not hold (Gujarati, 2003). If the p-value turns out to
be greater than the 5% level of significance, then the null hypothesis cannot be rejected, thus
the stochastic variables have a normal distribution. However, if the p-value will turn out to be
less than the 5% level of significance, the null hypothesis that there is a normal distribution will
have to be rejected and the alternative hypothesis will replace the null hypothesis.
Based on the graph and table for the test of normality of the model, the p-value is
0.9001. Since 0.9001>0.05, the null hypothesis cannot be rejected. The stochastic random
variable across the 75 observations does indeed follow a normal distribution.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
b) Test for Multicollinearity
Multicollinearity means the existence of either a perfect or imperfect linear relationship
among some or all explanatory variables of a regression model. It is important for one to detect
and cure the problem of multicollinearity since in the case of perfect multicollinearity, it will be
impossible for one to determine the coefficients of the X variables and their standard errors will
be infinite; in the case of imperfect multicollinearity, although the coefficients are determinate,
they possess large standard errors which makes the values of the coefficients inaccurate and
imprecise. (Gujarati, 2003)
Results for the test for multicollinearity with the use of auxiliary regressions can be seen from
the tables below.
Auxillary regression crime
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Auxillary regression literacy
Auxillary regression gdp
_cons 4.802848 14.11856 0.34 0.735 -23.3703 32.976 developed -3.42382 4.163973 -0.82 0.414 -11.7329 4.885259unemployment .032333 .1287346 0.25 0.802 -.2245529 .2892189corruption~x -1.037818 1.020595 -1.02 0.313 -3.074384 .9987481 gdp .0000384 .0001101 0.35 0.728 -.0001813 .0002581 literacy .0218052 .1446642 0.15 0.881 -.2668677 .3104781 poverty .2475159 .1150973 2.15 0.035 .0178428 .477189 crime Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 10943.8554 74 147.889937 Root MSE = 11.098 Adj R-squared = 0.1672 Residual 8375.50847 68 123.169242 R-squared = 0.2347 Model 2568.3469 6 428.057816 Prob > F = 0.0047 F( 6, 68) = 3.48 Source SS df MS Number of obs = 75
_cons 91.41476 4.167847 21.93 0.000 83.09795 99.73157 developed 1.924615 3.499487 0.55 0.584 -5.058504 8.907733unemployment -.0887703 .1074084 -0.83 0.411 -.3031004 .1255598corruption~x .8712675 .855371 1.02 0.312 -.8355989 2.578134 gdp .0000183 .0000923 0.20 0.843 -.0001659 .0002026 crime .0153174 .1016213 0.15 0.881 -.1874649 .2180996 poverty -.1148263 .0987157 -1.16 0.249 -.3118104 .0821579 literacy Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 7462.98315 74 100.851124 Root MSE = 9.3017 Adj R-squared = 0.1421 Residual 5883.48964 68 86.5219065 R-squared = 0.2116 Model 1579.49351 6 263.248918 Prob > F = 0.0107 F( 6, 68) = 3.04 Source SS df MS Number of obs = 75
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Auxillary regression corruption index
Auxillary regression unemployment rate
_cons -12017.47 15481.8 -0.78 0.440 -42910.93 18875.99 developed 16127.14 4169.131 3.87 0.000 7807.769 24446.51unemployment -17.98311 141.7173 -0.13 0.899 -300.7755 264.8093corruption~x 5199.177 939.7275 5.53 0.000 3323.98 7074.374 literacy 31.56794 159.1789 0.20 0.843 -286.0687 349.2046 crime 46.51138 133.3321 0.35 0.728 -219.5487 312.5715 poverty -90.75794 130.4336 -0.70 0.489 -351.0342 169.5183 gdp Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 3.3374e+10 74 451004194 Root MSE = 12213 Adj R-squared = 0.6693 Residual 1.0143e+10 68 149161702 R-squared = 0.6961 Model 2.3231e+10 6 3.8719e+09 Prob > F = 0.0000 F( 6, 68) = 25.96 Source SS df MS Number of obs = 75
_cons 2.188809 1.645106 1.33 0.188 -1.093948 5.471566 developed .7474914 .4850849 1.54 0.128 -.2204805 1.715463unemployment .0111903 .0151276 0.74 0.462 -.0189964 .041377 gdp .0000597 .0000108 5.53 0.000 .0000382 .0000812 literacy .0172487 .016934 1.02 0.312 -.0165426 .0510401 crime -.0144328 .0141933 -1.02 0.313 -.0427551 .0138895 poverty -.016072 .013891 -1.16 0.251 -.0437911 .0116471 corruption~x Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 345.779989 74 4.67270255 Root MSE = 1.3088 Adj R-squared = 0.6334 Residual 116.477158 68 1.71289938 R-squared = 0.6631 Model 229.302831 6 38.2171385 Prob > F = 0.0000 F( 6, 68) = 22.31 Source SS df MS Number of obs = 75
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Auxillary regression developed countries
_cons 7.934057 13.26997 0.60 0.552 -18.54576 34.41387 developed -1.010086 3.938177 -0.26 0.798 -8.868596 6.848423corruption~x .7133637 .9643602 0.74 0.462 -1.210987 2.637715 gdp -.0000132 .0001037 -0.13 0.899 -.0002202 .0001939 literacy -.1120319 .135554 -0.83 0.411 -.3825256 .1584618 crime .0286645 .1141282 0.25 0.802 -.1990749 .2564038 poverty .5007186 .0941064 5.32 0.000 .3129323 .6885049 unemployment Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 12425.5395 74 167.912695 Root MSE = 10.45 Adj R-squared = 0.3497 Residual 7425.21545 68 109.194345 R-squared = 0.4024 Model 5000.324 6 833.387334 Prob > F = 0.0000 F( 6, 68) = 7.63 Source SS df MS Number of obs = 75
_cons -.3364709 .4074573 -0.83 0.412 -1.149539 .4765976unemployment -.0009568 .0037306 -0.26 0.798 -.0084011 .0064874corruption~x .0451393 .0292932 1.54 0.128 -.0133143 .103593 gdp .0000112 2.89e-06 3.87 0.000 5.41e-06 .000017 literacy .0023009 .0041837 0.55 0.584 -.0060475 .0106493 crime -.0028753 .0034969 -0.82 0.414 -.0098533 .0041027 poverty .0009954 .0034449 0.29 0.774 -.0058788 .0078695 developed Coef. Std. Err. t P>|t| [95% Conf. Interval]
Total 15.5466667 74 .21009009 Root MSE = .32162 Adj R-squared = 0.5076 Residual 7.0337957 68 .103438172 R-squared = 0.5476 Model 8.51287096 6 1.41881183 Prob > F = 0.0000 F( 6, 68) = 13.72 Source SS df MS Number of obs = 75
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Aside from the auxilary regression, the VIF or Variance Inflation Factor could be use to test if
multicollineariy exist:
As a rule multicollinearity is said to be severe when the variance inflation factor is
greater than 10, while it is tolerable when the VIF is less than 10. We could see in the VIF table
that the mean VIF is equal to 2.01, denoting that 2.01 is less that 10, we could infer that the
regression model does not exhibit the problem of multicollinearity.
c) Test for Heteroscedasticity
Heteroscedasticity means that the variance of each disturbance term, conditional on the
chosen values of the explanatory variables, is not a constant number. This problem of
heteroscedasticity is a main violation of the assumption of equal variance. Thus, if one persists
in using the usual testing procedures despite this problem, whatever conclusion one draws may
be very misleading (Gujarati, 2003). It is thereby very important to find out whether
heteroscedasticity is present in the data, and if there is, to provide necessary measures. I then
used both the Breusch-Pagan test and the White test, as what have been discussed by Dr.
Rufino.
Mean VIF 2.01 unemployment 1.18 0.846368 crime 1.22 0.817365 literacy 1.24 0.804043 developed 2.21 0.452987corruption~x 2.91 0.343485 gdp 3.27 0.306080 Variable VIF 1/VIF
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Since the computed chi square of 6 is less than the critical chi square of 33.386031, and
the p-value of 0.090009 is greater than 0.05, then the null hypothesis that the disturbance
terms have constant variance must be accepted while the alternative hypothesis that the
stochastic error terms do not have constant variance must be rejected. There is no
heteroscedasticity in the model.
Another test that could be done to test the heteroscedasticity of the model is the white
test:
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Since the computed chi square of 26 is less than the critical chi square of 43.71306, and
the p-value of 0.096225 is greater than 0.05, then the null hypothesis that the disturbance
terms have constant variance must be accepted while the alternative hypothesis that the
stochastic error terms do not have constant variance must be rejected. There is no
heteroscedasticity in the model, just like in the Breusch-Pagan test.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
d) Test for Misspecification (Specification of Bias)
The Ramsey RESET Test was used for this case to determine whether there is a
specification bias in the model. The test revealed that there is no misspecification. With a
computed F equal to 0.0524347, this value is less than the critical F of 0.82, meaning that it is
insignificant at the 95 % confidence level. Consequently, the null hypothesis that there is no
specification bias in the model must be accepted.
e) Corrected Model and OLS Estimates
Since there were no diagnosed violations after the several tests were done. The
stochastic random variable indeed exhibits a normal distribution and is following
homoscedasticity. The model is also free from any prefect multicollinearity. The tests done
above were necessary for had they been diagnosed with violations, they would have provided
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
inaccurate and incorrect estimates of the parameters for the model. Thus, it would lead to
incorrect and invalid inferences and interpretations for the study. Here now are the final
corrected model and OLS estimates.
We should now test each exogenous variable to determine if is really significant to the
endogenous variable
f) Test for Significance of the OLS Estimates
Upon testing, we could see if each variable is significant on the OLS estimates.
Since the probability is less than the 5% level of significance, then the null hypothesis
that Crime does not affect Poverty is rejected and replaced with the alternative hypothesis that
the variable is actually significant to the model. The Crime variable is then statistically
significant.
Prob > F = 0.0351 F( 1, 68) = 4.62
( 1) crime = 0
CORRECTED MODEL FROM TESTING OLS REGRESSION:
Poverty = 36.01429 + .257699Crime - .1699038Literacy - .0000779Gdp – 1.201231Corindx + .5870588Unempl + 1.231924Develop + u
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Since the probability is greater than the 5% level of significance, then the null hypothesis
that Literacy does not affect Poverty is approved and the alternative hypothesis is then rejected
that the variable is actually not significant to the model. The Literacy variable is then not
statistically significant.
Since the probability is greater than the 5% level of significance, then the null hypothesis
that Gdp does not affect Poverty is approved and the alternative hypothesis is then rejected
that the variable is actually not significant to the model. The Literacy variable is then not
statistically significant.
Since the probability is greater than the 5% level of significance, then the null hypothesis
that Corruption does not affect Poverty is approved and the alternative hypothesis is then
rejected that the variable is actually not significant to the model. The Corruption variable is
then not statistically significant.
Prob > F = 0.2488 F( 1, 68) = 1.35
( 1) literacy = 0
Prob > F = 0.4889 F( 1, 68) = 0.48
( 1) gdp = 0
Prob > F = 0.2513 F( 1, 68) = 1.34
( 1) corruptionindex = 0
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Since the probability is less than the 5% level of significance, then the null hypothesis
that Unemployment does not affect Poverty is rejected and replaced with the alternative
hypothesis that the variable is actually significant to the model. The Unemployment variable is
then statistically significant.
Since the probability is greater than the 5% level of significance, then the null hypothesis
that Develop does not affect Poverty is approved and the alternative hypothesis is then
rejected that the variable is actually not significant to the model. The Develop variable is then
not statistically significant.
g) The Final Model and OLS Final Estimates
We can safely say that this is the final model and OLS estimates, we could now analyze
the significant variables to the model, we could retain the insignificant variables but they are
not needed to be analyze anymore.
Prob > F = 0.0000 F( 1, 68) = 28.31
( 1) unemployment = 0
Prob > F = 0.7735 F( 1, 68) = 0.08
( 1) developed = 0
FINAL MODEL FROM TESTING EXOGENOUS VARIABLES:
Poverty = 36.01429 + .257699Crime - .1699038Literacy - .0000779Gdp – 1.201231Corindx + .5870588Unempl + 1.231924Develop + u
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Based on the final model and OLS final estimates the Regressors of Crime and Unemployment,
the Regressors of Corruption index, developed countries, GDP per capita, and Literacy rate are
insignificant.
This is the Final analysis of significant variables to the final model and OLS estimates.
The coefficient of crime over population is 0.2572699, signifying that as total crime over
population increases by 1, population under poverty line will increase by
approximately .2572699 base points. The computed t is equal to 2.15, which is less than the
critical t. At the 95% confidence level, we could see that P>|t| is 0.035 which is less than 5%,
meaning that the total crime over population is significant to the endogenous variable which is
Poverty.
Now let us move on to the next exogenous variable which is unemployment rate, the
coefficient of unemployment rate is .5870588, signifying that as the unemployment rate of a
country’s population increases by 1, population under poverty line will increase by
approximately .5870588 base points. The computed t is equal to 5.32, which is really greater
than the critical t. At the 95% confidence level, we could see that P>|t| is 0.00001 which is
really less than 5%, meaning that the unemployment rate of a country’s population is very
significant to the endogenous variable which is Poverty.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
The researcher therefore concludes that poverty all over the world is very evident to our
surroundings in this present era. Poverty has affected every country out there either developed
or developing. It has been a major problem for different government to solve this problem. As a
student of De La Salle University – Manila, as I have research up different aspect and factors
that affects poverty it led me to different pictures and experiences of those people living below
the poverty line, they cannot even sustain their own needs to be able to survive and sustain
themselves. As a research and read different articles I saw many proposed solutions on how
this problem could be solved. It has been said in different articles that the crime rate of a
country affects poverty, they argue that it is the other way around but in a study made by
(Johnson, 2008). Crime rate of a country could later affect on the poverty line of their
respective country because it is like a cycle wherein as crimes either physical or blue-collar, this
could later on build up into a bigger mess in the future. In my study the total crime over
population is significant to the model.
Next is the literacy rate, the model showed that it is not significant because, literacy
rate is only based on the ability of a person to read and write, not the total educational
background of that country itself. Here in the Philippine although we have a literacy rate of
about 94%, we are living in a state of manner wherein many people are living below the
poverty line thus their basic earnings is not even high enough to reach minimum wage, that in
the direct effect they would not be able to sustain themselves. The literacy rate is really not
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
significant to the effects of it to poverty given this information. Next would be the GDP per
Capita, or the PPP which is that total services and goods made over population, this is also not
significant because of the fact that many factories and companies and using mechanical or
technological machines to accomplish their products, the human labor is rarely used nowadays.
So it does not mean that if the GDP per capita of a country is high, it has a low poverty line; thus
making GDP per capita insignificant to the model.
The next exogenous variable is the corruption index, it shows that it is insignificant to
poverty, maybe because of the fact that corruption is not really the problem but it is just the
excused of many to not work, they only depend on the government, who by itself is corrupted
in the inside. So we should just motivate ourselves to do our best so that we could be able to
strengthen and elevate ourselves in the poverty line by working hard.
Next is the unemployment rate which is significant to the model, unemployment rate is
significant because if less of the labor force is working, and if the people themselves doesn’t
work or if there are no job opportunities available. How could they be able to earn money to
sustain their daily needs? This topic and another problem itself is a significant variable that
clearly affects poverty.
As we move on to the different variables that affects poverty; we should be guided upon
that we are very fortunate that we are not living below the poverty line and we ourselves is
able to get the different needs and wants that we favor to get. Let us remember that majority
of the people doesn’t even have anything to eat. So let’s be responsible and humble to the
things that we have and by not wishing for anything more of what we really have.
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Bartle, Phil. "Factors of Poverty; The Big five." Home Page; Methods to Strengthen
Communities; Community Self Management, Empowerment and Development. N.p.,
n.d. Web. 12 Dec. 2011.Blanchard, O. (2006). Introduction to Macroeconomics.
Singapore: Pearson Education South Asia Ltd
Gujarati, D.M. (2003). Basic Econometrics. Singapore: McGraw-Hill Education (Asia)
Levine, R. (1992). A Sensitivity Analysis of Cross-Country Growth Regression. The
American Economic Review, 82, 942-963
Shah, A. (n.d.). Poverty Around The World — Global Issues. Global Issues : social,
political, economic and environmental issues that affect us all — Global Issues.
Retrieved December 12, 2011, from http://www.globalissues.org/article/4/poverty-
around-the-world#Introduction
Poverty And Corruption - Forbes.com. (n.d.). Information for the World's Business
Leaders - Forbes.com. Retrieved December 12, 2011, from
http://www.forbes.com/2009/01/22/corruption-poverty-development-biz-
corruption09-cx_mj_0122johnston.html
AfDB says fighting corruption and poverty in Africa go together - African Development
Bank. (n.d.). African Development Bank - Building today, a better Africa tomorrow.
Retrieved December 12, 2011, from
http://www.afdb.org/en/news-and-events/article/afdb-says-fighting-corruption-and-
poverty-in-africa-go-together-8633/
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Low literacy rate accounts for high poverty rate in Region - ModernGhana.com. (n.d.).
Ghana HomePage - Breaking News, Business, Sports, Entertainment and Video News.
Retrieved December 12, 2011, from
http://www.modernghana.com/news/180239/1/low-literacy-rate-accounts-for-high-
poverty-rate-i.html
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
A. Data Set (Tabulations of Data)
Countries Poverty Crime Literacy GDP Albania 12.5 3.29 99.1 3716
Argentina 30 5.24 97.6 9131 Armenia 26.5 2.53 99.7 2840 Austria 6 0.58 99 44988
Azerbaijan 11 2 99.5 6008 Belarus 27.1 5.56 99.7 5771 Belgium 15.2 1.83 99 42845 Bolivia 30.3 10.64 90.7 1900
Bulgaria 21.8 2.27 98.3 6356 Burma 32.7 15.58 89.9 742 Canada 9.4 1.67 99 46303
Chile 18.2 8.1 96.5 11827China 18.65 1.21 96.4 4382
Colombia 45.5 40.1 92.7 6360 Costa Rica 16 8.28 95.9 7701
Croatia 17 1.61 98.7 13776 Denmark 12.1 1.4 99 55986 Dominica 30 5.92 88 6632
El Salvador 30.7 51.83 82 3618 Estonia 19.7 6.26 99.8 14405 France 6.2 1.35 99 40704
Georgia 31 7.57 100 2629 Germany 15.5 0.8 99 40274 Greece 20 1.1 97.1 27311
Hungary 13.9 1.47 99.4 13024 India 25 2.77 75.04 1371
Ireland 5.5 1.95 99 46298 Italy 16.3 1.16 98.9 34059
Jamaica 14.8 59.5 86 4915 Japan 15.7 0.45 99 42783
Korea, South 15 2.3 99 20756 Kuwait 13.2 1.38 94.5 37009
Kyrgyzstan 40 7.78 99.3 843
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Latvia 28 4.38 99.8 10680 Lithuania 4 8.61 99.7 11046
Luxembourg 3 1.46 99 108952 Malaysia 5.1 2.31 91.9 8423 Maldives 16 2.62 97 6773
Malta 36.1 1.47 92.4 19707 Mauritius 8 3.75 87.4 7590
Mexico 18.2 11.59 92.8 9522 Moldova 26.3 3.7 97.3 1630 Morocco 15 0.4 55.6 2861
Nepal 24.7 2.24 68.2 557 Netherlands 10.5 1 99 46986 New Zealand 10 1.25 99 32163
Norway 7 0.64 99 84144 Oman 70 0.65 81.4 19405
Panama 28.6 13.28 93.4 7601 Papua New 37 15.1 57.8 1465
Peru 34.8 3.21 89.6 5205 Philippines 32.9 6.44 93.4 2123
Poland 17 1.21 99.3 12323 Portugal 18 1.16 94.9 21542
Qatar 13 1 93.1 74901 Romania 13.1 2.47 97.6 7542
Russia 13 14.18 99.5 10356 Saudi Arabia 8 0.85 85 16267 Seychelles 8.8 8.44 91.8 10617 Slovakia 21 1.74 99 16104 Slovenia 12.3 0.55 99.7 23648
South Africa 50 36.54 88 7274 Spain 19.8 0.91 97.9 30639
Sweden 14.2 0.89 99 49183 Thailand 9.6 5.9 94.1 4992 Tunisia 7.4 1.68 77.7 4199 Turkey 17.1 2.94 90.8 10309 Ukraine 35 5.36 99.7 3013
United Kingdom 14 1.57 99 36164 United States 12 5.22 99 46860
Uruguay 27.4 5.78 97.9 11998 Venezuela 37.9 47.21 95.2 10049
Yemen 45.2 3.97 58.9 1284 Zambia 86 22.51 70.6 1221
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Zimbabwe 80 34.29 91.2 594
Countries Corruption Index Unemployment Developed Albania 3.2 13.5 0
Argentina 2.9 7.9 0 Armenia 2.7 7.1 0 Austria 7.9 4.5 1
Azerbaijan 2.3 0.9 0 Belarus 2.4 1 0 Belgium 7.1 8.5 1 Bolivia 2.7 6.5 0
Bulgaria 3.8 9.2 0 Burma 1.4 5.7 0 Canada 8.7 8 1
Chile 6.7 8.7 0China 3.4 4.1 1
Colombia 3.7 11.8 0 Costa Rica 5.3 7.3 0
Croatia 4.1 17.6 0 Denmark 9.3 4.2 1 Dominica 5.9 23 0
El Salvador 3.4 7 0 Estonia 6.6 17.5 0 France 6.9 9.5 1
Georgia 4.1 16.4 0 Germany 8 7.4 1 Greece 3.8 12 1
Hungary 5.1 10.7 0 India 3.4 10.8 0
Ireland 8 13.7 1 Italy 4.3 8.4 1
Jamaica 3 12.9 0 Japan 7.7 5.1 1
Korea, South 5.5 3.3 1 Kuwait 4.1 2.2 0
Kyrgyzstan 1.9 18 0 Latvia 4.5 14.3 0
Lithuania 4.9 17.9 0 Luxembourg 8.2 5.5 1
Malaysia 4.5 3.5 0
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
Maldives 2.5 14.5 0 Malta 5.2 7 1
Mauritius 5.4 7.5 0 Mexico 3.3 5.6 0
Moldova 3.3 6.5 0 Morocco 3.3 9.8 0
Nepal 2.3 46 0 Netherlands 8.9 5.5 1 New Zealand 9.4 6.5 1
Norway 8.6 3.6 1 Oman 5.5 15 0
Panama 3.4 6.5 0 Papua New 2.1 1.8 0
Peru 3.7 6.7 0 Philippines 2.4 7.3 0
Poland 5 11.8 1 Portugal 5.8 10.7 1
Qatar 7 0.5 1 Romania 3.8 8.2 0
Russia 2.2 7.6 0 Saudi Arabia 4.3 10.8 0 Seychelles 4.8 2 0 Slovakia 4.5 13.5 0 Slovenia 6.6 10.6 0
South Africa 4.7 23.3 0 Spain 6.1 20 0
Sweden 9.2 8.3 0 Thailand 3.4 1.2 0 Tunisia 4.2 14 0 Turkey 4.4 12.4 0 Ukraine 2.2 8.4 0
United Kingdom 7.7 7.9 1 United States 7.5 9.7 1
Uruguay 6.7 6.8 0 Venezuela 1.9 12.1 0
Yemen 2.1 35 0 Zambia 3 50 0
Zimbabwe 2.2 95 0
B. Detailed Summarization of Variables
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
99% 86 86 Kurtosis 7.70871195% 50 80 Skewness 1.94985890% 37.9 70 Variance 251.753575% 30 50 Largest Std. Dev. 15.8667450% 17 Mean 22.09933
25% 12.3 5.5 Sum of Wgt. 7510% 7.4 5.1 Obs 75 5% 5.5 4 1% 3 3 Percentiles Smallest Poverty
99% 59.5 59.5 Kurtosis 10.1848795% 40.1 51.83 Skewness 2.77512890% 15.58 47.21 Variance 147.889975% 7.57 40.1 Largest Std. Dev. 12.16150% 2.53 Mean 7.412667
25% 1.35 .58 Sum of Wgt. 7510% .85 .55 Obs 75 5% .58 .45 1% .4 .4 Percentiles Smallest Crime
99% 100 100 Kurtosis 7.75538895% 99.7 99.8 Skewness -2.25293390% 99.7 99.8 Variance 100.851175% 99 99.7 Largest Std. Dev. 10.0424750% 97.6 Mean 93.04453
25% 91.2 68.2 Sum of Wgt. 7510% 81.4 58.9 Obs 75 5% 68.2 57.8 1% 55.6 55.6 Percentiles Smallest Literacy
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
99% 108952 108952 Kurtosis 6.9679595% 55986 84144 Skewness 1.87883690% 46303 74901 Variance 4.51e+0875% 30639 55986 Largest Std. Dev. 21236.8650% 10309 Mean 18735.67
25% 4382 843 Sum of Wgt. 7510% 1465 742 Obs 75 5% 843 594 1% 557 557 Percentiles Smallest GDP
99% 9.4 9.4 Kurtosis 2.2086995% 8.9 9.3 Skewness .537989390% 8 9.2 Variance 4.67270375% 6.6 8.9 Largest Std. Dev. 2.16164350% 4.3 Mean 4.8
25% 3.2 2.1 Sum of Wgt. 7510% 2.2 1.9 Obs 75 5% 2.1 1.9 1% 1.4 1.4 Percentiles Smallest Corruption Index
99% 95 95 Kurtosis 25.3706995% 35 50 Skewness 4.24666190% 18 46 Variance 167.912775% 13.5 35 Largest Std. Dev. 12.9581150% 8.4 Mean 11.66933
25% 6.5 1.2 Sum of Wgt. 7510% 3.3 1 Obs 75 5% 1.2 .9 1% .5 .5 Percentiles Smallest Unemployment
December 11, 2011
Dr. Cesar Rufino - ECONMET Page
99% 1 1 Kurtosis 1.82418595% 1 1 Skewness .907846590% 1 1 Variance .210090175% 1 1 Largest Std. Dev. .458355950% 0 Mean .2933333
25% 0 0 Sum of Wgt. 7510% 0 0 Obs 75 5% 0 0 1% 0 0 Percentiles Smallest Developed