Content uploaded by Moses Kwadzo
Author content
All content in this area was uploaded by Moses Kwadzo on Sep 06, 2022
Content may be subject to copyright.
This article was downloaded by: [Moses Kwadzo]
On: 10 July 2015, At: 12:56
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: 5 Howick Place, London, SW1P 1WG
Click for updates
Journal of Poverty
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/wpov20
Choosing Concepts and Measurements of
Poverty: A Comparison of Three Major
Poverty Approaches
Moses Kwadzoa
a Department of Agricultural Economics and Extension, University of
Cape Coast, Cape Coast, Ghana
Published online: 09 Jul 2015.
To cite this article: Moses Kwadzo (2015): Choosing Concepts and Measurements of
Poverty: A Comparison of Three Major Poverty Approaches, Journal of Poverty, DOI:
10.1080/10875549.2015.1015067
To link to this article: http://dx.doi.org/10.1080/10875549.2015.1015067
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Journal of Poverty, 00:1–15, 2015
Copyright © Taylor & Francis Group, LLC
ISSN: 1087-5549 print/1540-7608 online
DOI: 10.1080/10875549.2015.1015067
Choosing Concepts and Measurements
of Poverty: A Comparison of Three Major
Poverty Approaches
MOSES KWADZO
Department of Agricultural Economics and Extension, University of Cape Coast,
Cape Coast, Ghana
A comparative analysis of different poverty measures, particularly
across studies that use different conceptualizations and measure-
ments of poverty, is very valuable. In highlighting this fact, this
article compares three poverty measurements: monetary poverty,
social exclusion, and capability poverty measurements. The results
indicate that all three poverty measurements classify varied pro-
portions of the U.S. population as poor. These variations occur as
a function of the conceptualization and measurement of poverty.
In general, all three poverty measurements are inadequate indica-
tors of well-being. It is reasonable to suggest that researchers report
results using more than one poverty measurement.
KEYWORDS social exclusion poverty, monetary poverty, capabil-
ity poverty, measurement of poverty, conceptualization of poverty
INTRODUCTION
Poverty and social policy research utilizes different poverty definitions and
measurements. However, different poverty definitions and measurements
have been associated with varied and/or contradictory evaluation outcomes
(Bell, 1995; Laderchi, Saith, & Stewart, 2003). Consequently, the choice of a
particular poverty definition and measurement has important consequences
for the poor. According to Laderchi et al. (2003), the various conceptualiza-
tions and measurements of poverty may not indicate that the same people are
poor. Hagenaars and Vos (1988) also emphasized that the choice of a specific
Address correspondence to Moses Kwadzo, Department of Agricultural Economics and
Extension, University of Cape Coast, Cape Coast, Ghana. E-mail: Kwamo50@gmail.com
1
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
2M. Kwadzo
definition and measurement of poverty may result in different estimates of
a population as poor or as not poor. Interestingly, researchers and policy-
makers often prefer to adopt a particular definition of poverty based on their
specific study objective, political interest, the availability of data, and other
forms of justification (Behn, 1995,2003).
In the United States, poverty policy and program evaluations are linked
to different estimates of the numbers of poor or not poor. The difference
in poverty estimates could be attributed to differences in the conceptualiza-
tion and measurement of poverty rather than to the facts on the ground.
In that context, many poverty studies use indicators of monetary poverty,
whereas others employ indicators of either capability or social exclusion
poverty to support their findings and conclusions (Blank, 2007; Hagenaars &
Vos, 1988; Laderchi et al., 2003;Rank,2004). Although the choice of a spe-
cific poverty measure may have major consequences for poverty reduction,
some measures may identify specific poverty situations better than others
(Hagenaars & Vos, 1988; Laderchi et al., 2003). What is lacking in previ-
ous studies is no comparative analysis of different poverty indicators such as
monetary poverty, capability poverty, and social exclusion poverty. This lack
of comparative analysis makes it difficult to conclude which poverty mea-
surement is more appropriate in a given poverty situation. Which poverty
indicator(s) should be used to define and measure poverty? This article aims
to compare and examine how the different poverty measurements (mone-
tary, capability, and social exclusion poverty measurements) estimate poverty
outcomes.
CONCEPTS AND MEASUREMENTS OF POVERTY
Poverty is a worldwide concern. Although there is global concern about
poverty reduction, there is no single definition and measurement of poverty
(Kotler, Robert, & Leisner, 2006; Laderchi et al., 2003). Because poverty
affects heterogeneous groups, the concept of poverty is relative depend-
ing on different interest groups and the individuals experiencing poverty
(Kotler et al., 2006;Rank,2004). Laderchi et al. (2003) underscored four
main approaches to poverty definition and measurement: monetary poverty,
capability poverty, social exclusion poverty, and participatory poverty
approaches. This article focuses on the first three approaches. Unlike the
other three poverty measurement approaches, the participatory approach
emphasizes the need to get the poor themselves involved in defining what it
means to be poor. The participatory approach employs qualitative methods
and tools including participatory rural appraisal. Because it is difficult for
the researcher to purposively sample participants from different parts of the
nation to define what it means to be poor, this approach was not included
in this study (Laderchi et al., 2003).
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 3
Monetary Poverty
The monetary approach is a commodity- or utility-based approach that
defines poverty in the context of the distribution and utilization of goods as
well as the fact of individuals possessing such goods. The monetary approach
uses methodologies that emphasize monetary income based on the assump-
tion that all relevant heterogeneity between individuals can be controlled
for and that income or expenditure indicators are appropriate proxies for
consumption. Because the monetary poverty approach focuses on data from
market-purchased essential goods and this reflects the utility levels of such
goods, it often analyzes income or consumption data (Laderchi et al., 2003;
Ravallion, 1998). The approach employs an absolute income threshold com-
puted by putting a monetary value on the minimum amount of goods a
family or individual needs to survive. When a family’s or an individual’s total
income falls below the poverty threshold, the family or individual is consid-
ered to be poor. Thus, monetary poverty is defined as a shortfall in family or
personal income (expenditure) below some minimum level of resources, a
poverty line (Laderchi et al., 2003). In other words, monetary poverty is hav-
ing less than an objectively defined (absolute minimum) amount of resources
in society. Although monetary poverty is useful for interpersonal comparison
in poverty analysis, it fails to take into account variations in personal char-
acteristics as well as disparities in social environment (Laderchi et al., 2003;
Sen, 1985).
There are two slightly different versions of the U.S. poverty mea-
surements, the federal poverty thresholds and poverty guidelines (U.S.
Department of Health and Human Services, 2007,2009). The poverty thresh-
old is a set of money income thresholds that vary by family size and
composition (based on the ages and number of family members). The
poverty guidelines are a simplified version (percentage multiples such as
150% or 1.5) of the federal poverty thresholds that are used for administrative
purposes by many government aid programs to determine an individual’s eli-
gibility for certain federal programs (U.S. Department of Health and Human
Services, 2007,2009). According to the U.S. Census definition, an individual
or family is considered to be poor if the annual before-tax money income is
less than the poverty threshold (U.S. Bureau of Census, 2004).
Social Exclusion Poverty
Social exclusion poverty is the relative deprivation of a person or the per-
son’s lack of access to certain commodities or services common to others in
society (Laderchi et al., 2003). The concept of social exclusion was advanced
in industrialized countries to demonstrate the process of marginalization and
deprivation. The origin of social exclusion can be traced back to a break-
down in social cohesion that followed civil unrest in France in 1960 resulting
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
4M. Kwadzo
from increasing unemployment and socioeconomic inequalities. The use of
the concept of social exclusion spread from France to other European nations
and then to the rest of the world. According to Mathieson et al. (2008)and
Laderchi et al. (2003), the concept of social exclusion has been subjected
to different meanings over the years, such that it has no single acceptable
definition. The term has also been used to emphasize an individual’s lack of
participation in society as well as lack of access to citizenship rights. Social
exclusion is also used to describe groups that are at risk of exclusion and
the states associated with exclusion.
Mathieson et al. (2008), Laderchi et al. (2003), and Silver and Miller
(2003) outlined attributes or key elements of social exclusion as being mul-
tidimensional, dynamic, and relational. The multidimensionality of social
exclusion relates to the fact that the experience of exclusion exists in many
spheres such as economic, social, and political. The dynamic process of
social exclusion posits that one form of exclusion can lead to more exclusion
and permanent multiple forms. The relationship aspect of social exclusion
relates to the fact that social exclusion occurs as a result of unequal social
relationships or differential power. As a process, social exclusion maintains
unequal power relationships, which create inequality.
Mathieson et al. (2008) noted that the lack of a single definition of social
exclusion has left the concept with no single validated exclusion poverty
measurement. Although the UNDP’s (1997) Human Development Index has
been used to describe some aspects of social exclusion, there are no specific
indicators of social exclusion at the global level. However, two sets of indica-
tors (primary and secondary indicators) have been adopted (Department for
Work and Pensions, 2003). The primary indicators include persistent at-risk
poverty rate, relative median poverty gap, long-term unemployment rate,
and population living in jobless households. The primary indicators are at-
risk poverty rate, poverty risk by household type, and poverty risk by work
intensity of household. The primary indicators cover the most important
aspects of social exclusion, whereas the secondary indicators give further
clarification to the primary indicators.
In contrast to using absolute income poverty, the European Union (EU),
researchers, and governments have adopted relative income poverty as the
dominant standard measure of social exclusion poverty to capture the inci-
dence of poverty. Specifically, the EU measures the risk of poverty at 60%
of median income. Fifty percent of the median income is also considered as
an indicator of relative income poverty. The median income threshold used
by the EU is assumed to express important information about the nature of
deprivation in a specific society (Hilamo, Saillila, & Sund, 2004). Unlike the
absolute poverty threshold, the relative income measure takes into account
the variations in personal characteristics as well as disparities in the social
environment.
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 5
Capability Poverty
Capability poverty is the deprivation of a person or the failure of a person
to develop capabilities to achieve a certain level of functioning. Capability
refers to the individual’s abilities (a list of commodities he or she possesses
in a given external environment) to adequately fulfill certain crucial roles at
a minimum, whereas functioning refers to what a person has succeeded in
achieving or doing with his or her abilities (Laderchi et al., 2003; Saith, 2001;
Sen, 1985). The capability poverty approach argues that monetary resources
are a necessary condition and that other resources are required to develop
a person’s capabilities to prevent the occurrence of poverty (Laderchi et al.,
2003). Nussbaum’s (2000) basic human functional capabilities provide a uni-
versal list that allows for further negotiation. Items on Nussbaum’s capability
list include, but are not limited to, physical health, a person’s control over his
or her environment, his or her institutional or other affiliations, and his or her
emotions. Limited capability (i.e., basic requirements for a decent life) has
been described as including health, nutrition, education, sanitation, water
supply, and housing (Laderchi et al., 2003).
Given the multidimensionality of capability poverty, a comparison of
the overall capability set is more difficult to examine (Laderchi et al., 2003).
According to Laderchi et al. (2003), selecting the capability set or setting a
capability poverty threshold is somewhat arbitrary, as this is based on the
general standard of a given region or country. The UNDP (1997) defined
human poverty as deprivation in three essential elements including life
expectancy, knowledge, and a decent standard of living. Human poverty
in developing nations has been defined in terms of the following indicators
in each nation: having a life expectancy of fewer than 40 years at birth, the
prevalence of under-5 year mortality, adult illiteracy, and a general lack of
access to improved water sources. The UNDP (2001) adopted a different
poverty index for the various developed nations with life expectancy set at
below 60 years and with literacy defined as lack of functional literacy among
adults (Laderchi et al., 2003).
Laderchi et al. (2003) pointed out that the assessment and comparison
of an interpersonal capability set may involve the assessment of just one
element considered as representative of the set. In this way, the value of the
capability set is equated with that of a single element of the set, such as being
educated (Laderchi et al., 2003). For instance, if the chosen indicator is the
years of education, the comparison between individuals would relate to their
levels of achieved education, for example, education until Grade 6. Many
empirical studies on capability poverty employ one or more indicators of
education, health, nutrition, sanitation, water supply, and housing (Darling,
2002; Laderchi et al., 2003).
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
6M. Kwadzo
METHOD
Data
This research utilized data from the 2004 National Longitudinal Study of
Youth, 1979 (NLSY79) from the U.S. Department of Labor (2006). The
NLSY1979 is a national representative sample of 12,686 young women and
men between 14 and 22 years of age when they were first interviewed in
1979. The 2004 survey was the 21st wave of the NLSY79 panel data. The age
distribution of the survey respondents ranged from 39 to 48 years. Data from
2004 were selected because the data set contains observations on multiple
phenomena that provide data for the analytical basis of this article. In 2004,
the retention rate for the NLSY79 respondents was 76.9% (N=7,661). The
2004 interview was conducted between January 2004 and February 2005. The
socioeconomic characteristics of the survey sample and the national popula-
tion based on the 2004 American Community Survey (ACS) were compared
with the extent to which the sample matched well with the national pop-
ulation of the United States. The comparison shows that the sample survey
was generally representative of the national population in terms of socioe-
conomic characteristics including sex, educational achievement, family size,
and family income (USBC, 2007).
Methodological Approach and Poverty Estimation
As stated earlier, this article focuses on monetary poverty, capability poverty,
and social exclusion poverty approaches. To estimate the size of the survey
population in poverty with respect to each approach, the researchers fol-
lowed two main steps. Step 1 involves the selection of variables by which
the resources of persons/households are to be measured. In Step 2, the
poverty lines below which people are considered to be poor are determined.
To identify the size of the population considered to be poor, the distribution
of the various variables was transformed, with those who are not poor being
coded 0 and those who are poor being coded 1.
Cross tabulation is then used to calculate the sensitivity and specificity
of the poverty measurements. The sensitivity of a model is the percentage of
the group that has the characteristics of interest (i.e., individuals in poverty)
which has been identified accurately (Pallant, 2007). The model specificity
is the percentage of the group without the characteristics of interest (that
is individuals not in poverty) that has been identified correctly (Pallant,
2007). Kappa measure of agreement statistics are used to assess inter-rater
agreement or the consistency of two different events.
Variable Categorization
In this article, monetary poverty is measured at two levels. The first poverty
line is the U.S. federal poverty threshold. The second poverty measurement
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 7
is the U.S. poverty guideline, which is 1.5 times the federal poverty threshold
(U.S. Department of Health and Human Services, 2007,2009). In this regard,
individuals or households with a total income less than either the threshold
or the guideline are considered to be poor. Individuals or households in
poverty were coded 1, and those not in poverty were coded 0.
In this study, social exclusion poverty is operationalized in terms of
household median income. Social exclusion poverty is measured on two
levels. Using a narrow poverty level, households with a total income below
50% of the household median income of the population age between 39 and
48 are considered to be poor. In addition, using a broader poverty level,
households with a total income below 60% of the median income are con-
sidered to be poor. The corresponding weighted 2004 household median
national income for the population age between 39 and 48 years was
$60,023.00 (U.S. Bureau of Census, 2005).
Unlike the monetary and social exclusion approaches, capability poverty
measurement utilizes a non-income-based approach. In that context, poverty
is defined with respect to educational achievement or how well an individual
is educated (Saith, 2001). Education is chosen as an indicator of capabil-
ity poverty because, in industrialized countries, there is a higher likelihood
that the average person will have a right to 10 or more years of education.
Education is considered as one of the most critical and commonly agreed
capabilities (Hulme & McKay, 2005). Consequently, the capability poverty
threshold is defined as the end of specific years of schooling. First, the
educational threshold is identified by matching total income to educational
level. The threshold is set at a point at which income falls below the federal
poverty line. A high poverty rate matches an educational level less than a
high school diploma (USBC, 2008). Because high-income poverty tends to
concentrate at the lowest educational level, the authors restricted the edu-
cation poverty thresholds to high school diploma or less. The authors used
two levels of measurement with respect to capability poverty. First, individ-
uals whose educational attainment is 11 years of schooling or less (Grade
11) are considered to be poor. Second, individuals whose education attain-
ment is 12 years of schooling or less (Grade 12) are considered to be poor.
Capability poverty was measured by asking about the highest grade com-
pleted by the respondents. The responses ranged from no schooling, first
grade to eighth year of college, or more. The distribution of the respondents’
highest grade completed was transformed into a dummy variable at two dif-
ferent levels. First, capability poverty is a dummy variable scored 1 if the
respondent indicated education level to be Grade 11 or less; all others are
0. Second, capability poverty is a dummy variable scored 1 if the respon-
dent indicated the highest grade completed to be Grade 12 or less; all others
are 0.
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
8M. Kwadzo
TABLE 1 Percent Distribution of Sample in Poverty
Type of poverty Measured
%in
sample
Monetary poverty Poverty status at federal poverty threshold cutoff point 14.5
Poverty status at 1.5th cutoff point 21.4
Social exclusion poverty Household income at 50% cutoff point of the median
income of the population age between 39 and 48 years
30.6
Household income at 60% cutoff point of the median
income of the population age between 38 and 49 years
36.7
Capability poverty Equal to or less than Grade 11 cutoff point 10.6
Equal to or Grade 12 cutoff point 54.2
RESULTS
Distribution of Sample in Poverty
Table 1 shows the percentage distribution of the sample in poverty. It is
obvious that the distribution of the sample below the thresholds of the three
poverty measurements increases with increasing cutoff points. When the offi-
cial federal poverty threshold was applied, 14.5% of the sample fell below
the monetary poverty threshold. When the federal poverty threshold was set
at 1.5 times the official federal poverty threshold, 21.4% of the sample fell
below the monetary poverty threshold. In contrast, social exclusion captured
more people in poverty than monetary poverty. With the social exclusion
poverty threshold set at 50% of median national income of the population
aged between 39 and 48, 30.6% of the sample fell below the poverty thresh-
old. At the 60% median income cutoff point, 36.7% of the sample fell below
the social exclusion poverty threshold. With the capability poverty cutoff
point equal to or less than the Grade 11 cutoff point, 10.6% of the sample
was below the capability poverty threshold. This figure is less than the per-
centage in monetary or social exclusion poverty at their respective first cutoff
levels. At the cutoff point equal to or less than the Grade 12 cutoff point,
however, 54.2% of the sample was in capability poverty.
Rate of Poverty Distribution
To obtain a bigger picture of how poverty distribution rates are influenced
when the three poverty measurement threshold levels are increased, cumu-
lative frequency curves were drawn for the three poverty measurements,
as shown in Figures 1 through 3. The monetary poverty curve is almost
similar to that of the social exclusion poverty curve. There is a gradual
change (increase) in the distribution of the sample in poverty as the cut-
off points of the monetary and social exclusion poverty thresholds increase.
With the capability poverty measurement, there is not much change in the
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 9
Monetary Poverty
0
10
20
30
40
50
60
70
80
90
1 1.2 1.5 2 3 4 5 6
Levels of Official Poverty Threshold
Percent in Poverty
Series1
FIGURE 1 Distribution of people in monetary poverty.
Social Exclusion Poverty
0
10
20
30
40
50
60
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Household’s Proportion of Median
National Income of the Population
Aged Between 39 and 48
Percent in Poverty
Social Exclusion
Poverty
FIGURE 2 Distribution of people in social exclusion poverty.
distribution of the sample in poverty until the highest educational grade
completed reaches Grade 12.
Overlap Between Poverty Measures
The authors tested the degree of agreement between the different measure-
ments of poverty. The goal is to see whether individuals identified as falling
below the poverty line were also classified as falling below the poverty line
by the other measures. A kappa value of 0.5 represents moderate agreement,
above 0.7 represents good agreement, and above 0.8 represents very good
agreement (Peat, 2001, p. 228).
A cross tabulation of the monetary poverty measurement (at the federal
poverty threshold cutoff point) and the social exclusion poverty measure-
ment at 50% of the median national income threshold cutoff point indicates
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
10 M. Kwadzo
Capability Poverty
0
20
40
60
80
100
120
1 3 5 7 9 1113151719
Highest Grade Completed
Percent in Sample
Series1
FIGURE 3 Distribution of people in capability poverty.
that, out of 2026 individual respondents classified as being in social exclusion
poverty, 958 are also identified under the official federal poverty threshold.
A sensitivity value of 47.3% (958/2026) represents the proportion of those
individuals classified under the monetary and the social exclusion measures
who are correctly identified.
A cross tabulation of the capability poverty measurement (when
an individual’s educational achievement is equal to or less than Grade
11) and the monetary poverty measurement (at the federal poverty thresh-
old cutoff point) shows a sensitivity value of 38.7 (258/667). A sensitivity
value of 38.7 (258/667) implies that 38.7% of those individuals classi-
fied under the monetary and the capability poverty measurements are
correctly identified. A kappa measure of agreement value (kappa =.23,
p<.001) indicates a weak consistency between the two poverty measures.
A cross tabulation of the capability poverty measurement (when an indi-
vidual’s educational achievement is equal to or less than Grade 11) and
social exclusion poverty (at 50% of the median national income of the pop-
ulation age between 39 and 48 years) indicated a sensitivity value of 64.5
(430/667). The sensitivity value of 64.5 (430/667) indicates that 64.5% of
those individuals classified under the social exclusion and the capability
poverty measurements are correctly identified. A kappa measure of agree-
ment value (kappa =.2, p<.001) indicates a weak consistency between the
two poverty measurements.
DISCUSSION
The majority of poverty studies in the United States employ the federal
poverty threshold (a monetary poverty measurement), whereas a few use
indicators of capability and social exclusion poverty measures. The U.S.
federal poverty measure has been criticized as a less effective poverty
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 11
measurement (Blank, 2007; Dalaker, 2005; National Poverty Center, 2006;
Rank, 2004). In particular, the use of different poverty measurements in the
U.S. poverty program evaluations has been associated with different and/or
contradictory outcomes. In general, this study did not directly address these
controversies; instead, it informs the literature on poverty measurement in a
number of directions. Laderchi et al. (2003) noted that the use of different
poverty measurements would likely result in different numbers of people
being classified as in poverty. The variability in the levels of the percentage
distribution of an individual crossing the poverty line for the three poverty
measurements can be attributed to differences in the theoretical assumptions
of these poverty measurements (Laderchi et al., 2003; World Bank, 2002). The
diversity of the assumptions as well as the methodology adopted for each
poverty measurement influence the extent to which each poverty measure-
ment is able to capture a particular individual as being poor. The assumption
of the monetary poverty measure is based on absolute poverty and focuses
on the income that classifies a family or individual as poor when income or
consumption falls below an objectively derived poverty line.
The percentage distribution of individuals in poverty for the three
poverty measurements indicates that poverty incidence is sensitive to change
when any of the poverty thresholds is increased. Monetary and social exclu-
sion poverty measurements exhibit a consistent pattern of gradual increase
in the percentage distribution of the sample in poverty when the threshold
levels are raised. This observation could be attributed to the fact that the
monetary and the social exclusion poverty approaches utilize income-based
indications. Social exclusion poverty has the highest percentage distribution
of the sample in poverty. After social exclusion poverty, monetary poverty
comes next with a consistent count of the sample percentage in poverty
when the official federal poverty threshold level is raised. This disparity
may be explained by the fact that the absolute income measure of mon-
etary poverty does not consider differences in personal characteristics or
social environment. However, the social exclusion poverty measurement,
which employs the relative income measure, takes into account the variations
in personal characteristics as well as disparities in the social environment.
In particular, when the poverty threshold of educational achievement (capa-
bility poverty measure) was raised from equal to or less than Grade 11 to
equal to or less than Grade 12, the percentage in poverty was dramatically
increased (from to 10.6% – 54.2%). This sharp increase in the distribution
of people in capability poverty can partly be attributed to the institutional
arrangement of education in the United States, where elementary and high
school education is provided and enforced by the public sector, with control
and funding coming from federal, state, and local governments. Schooling is
compulsory for all children in the United States, but the age range for which
school attendance is required varies from state to state. Some states allow stu-
dents to leave school between age 14 and 17 years with parental approval
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
12 M. Kwadzo
(Gatto, 2003). It seems that compulsory and free education at elementary
and high school levels has promoted higher attendances.
Furthermore, the results of the cross tabulations indicated that the
poverty measurements overlap to capture a percentage of the sample as
being poor. In general, there are overlaps between the poverty measure-
ments to identify individuals who experience more than one type of poverty.
These overlaps range from about 39% to 65%. The kappa measure of agree-
ment value of 0.6 indicates a moderate consistency between the monetary
and social exclusion poverty measurements. Social exclusion poverty and
monetary poverty measurements have shown a consistent pattern of identi-
fying poverty in the sample. The kappa measure of agreement value between
capability poverty and monetary or social exclusion poverty indicates weak
agreement between the classifications of cases as falling below the poverty
lines. In that sense, the findings that the capability poverty measurement
exhibits inconsistent patterns of poverty distribution as well as a weak kappa
measure of agreement value with the poverty measures implies that the
capability poverty measurement is not a good poverty measurement (com-
pared with the other measures), particularly in a developed nation such as
the United States. One could also argue that part of the inconsistent mea-
surement by capability poverty may be a consequence of its nonincome
indicator (education) relative to the income indicators used for the other
poverty measurements.
The different poverty measurements result in different estimates of pop-
ulation sizes for those considered to be poor. Although the three poverty
measurements are different, they do overlap to capture a proportion of
the population as being poor. What distinguishes one poverty measurement
from another is the unique poverty definition and defined threshold levels
of each measure. They also result in information and conclusion discrepan-
cies, which in turn may have serious policy implications. Because monetary
poverty emphasizes the distribution and utility of goods as well as an indi-
vidual’s possession of such goods, poverty reduction policies often tend to
focus on economic growth and money income generation. For example, to
reduce poverty, monetary poverty policy emphasizes the requirement for an
increase in productivity and income as a means to reduce poverty among
the poor (Laderchi et al., 2003; Sen, 1985). It therefore focuses on policies
that promote economic growth, redistribution, and socioeconomic structure
changes that are fundamentally important to poverty reduction. Social exclu-
sion poverty policies tend to promote policies that minimize racial, gender,
and class discrimination and inequality to increase access to more resources
(Laderchi et al., 2003).
Alkire (2007) and Darling (2002) emphasize human capital or the capa-
bility to provide people with the necessary skills or abilities to function well
in society. According to them, the lack of human capability places an individ-
ual in a more economically vulnerable position when they face detrimental
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 13
events such as loss of job, illness, or family changes. Individuals with a high
quality of human capital will do well in the labor market. Schiller (2008)also
pointed out that personal investment in human capital such as education can
operate as a strong mechanism that can influence the distribution of poverty.
Policies of capability poverty alleviation emphasize increasing the accessibil-
ity of goods and services such as good health, nutrition, and education to
individuals or groups.
Indeed, it can be argued that all three poverty approaches discussed
in this article are inadequate indicators of well-being. As Rank (2004,
p. 21) aptly noted, poverty and its measurement involve a degree of subjec-
tivity. In his view, despite the importance of well-grounded thinking, poverty
is in the eye of the beholder. In this sense, choosing a particular poverty
approach and its corresponding poverty line(s) may give an appearance that
a certain number of people are in poverty and policies are generated based
on that. This hardly gives policymakers an accurate picture of poverty and
whether policies to eradicate it are really working. This means that policy-
makers should not focus their efforts on people living just below the poverty
line to give a rosy picture of the evolution of poverty and its eradication.
CONCLUSION
Current research reveals the complexity of poverty in terms of the linkage
between the theory, definition, and measurement of poverty. The different
poverty measurements result in different estimates of the studied sample
sizes considered to be poor. Although the three poverty measurements are
different, they do overlap to capture a segment of the population that is
poor. This variation occurs as a function of the definition and measure-
ment of poverty. Therefore, differences in conclusions among studies and
discrepancies among evidence for policymakers may, among other things,
stem from the variation in poverty definitions and means of measurement.
It is reasonable to suggest, therefore, that researchers seriously consider
reporting results using more than one measurement on account of the
multidimensional nature of poverty.
REFERENCES
Alkire, S. (2007). Choosing dimensions: the capability approach and multidimen-
sional poverty (CPRC Working Paper, 88). Retrieved from http://mpra.ub.uni-
muenchen.de/8862/1/MPRA_paper_8862.pdf
Behn, R. (1995). The big question of public management. Public Administration
Review,55(4), 313–324.
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
14 M. Kwadzo
Behn, R. (2003). Why measure performance? Different purposes require different
measures. Public Administration Review,63(5), 586–603.
Bell, C. S. (1995). What is poverty? American Journal of Economics and Sociology,
54(2), 161–162.
Blank, R. (2007). What we know, what we don’t know, and what we need to know
about welfare reform. National Poverty Center Working Paper Series,7(19), 1–8.
Dalaker, J. (2005). Alternative poverty estimates in the United States. Retrieved from
www.census.gov/hhes/www/p60-226sa.pdf
Darling, R. (2002). Partnership models in human services. Sociological foundations
and practices. New York, NY: Kluwer Academic/Plenum.
Department for Work and Pensions. (2003). United Kingdom National Action Plan
on Social Inclusion 2003–05: The annexes. London, England: Author. Retrieved
from http://www.dwp.gov.uk/publications/dwp/2003/nap/
Gatto, J. T. (2003). The underground history of American education.NewYork,NY:
Oxford Village Press.
Hagenaars, A., & Vos, K. (1988). The definition and measurement of poverty. Journal
of Human Resources,23(2), 211–221.
Hilamo, H., Saillila, S., & Sund, R. (2004). Luxembourg income study. Rethinking
relative measure of poverty. Retrieved from http://www.lisdatacenter.org/wps/
liswps/368.pdf
Hulme, D., & McKay, A. (2005). Identifying and understanding chronic poverty:
Beyond income measures. Manchester, England: Chronic Poverty Research
Center.
Jordan, B. (1996). A theory of poverty and social exclusion. Cambridge, MA: Polity
Press.
Kotler, P., Roberto, N., & Leisner, T. (2006). Alleviating poverty: A macro/micro
marketing perspective. Journal of Macromarketing,26(3), 233–238.
Laderchi, C., Saith, R., & Stewart, F. (2003). Does it matter that we do not agree on the
definition of poverty: A comparison of four approaches. Oxford Development
Studies,31(3), 233–274.
Mathieson, J., Popay, J., Enoch, E., Escorel, S., Hernandez, M., Johnson, H., &
Rispel, L. (2008). Social exclusion: Meaning, measurement and experience
and the link to health inequalities. Retrieved from http://www.who.int/social_
determinants/media/sekn_meaning_measurement_experience_2008.pdf.pdf
National Poverty Center. (2006). Poverty facts: Poverty in the United States. Retrieved
from http://npc.umich.edu/poverty/
Nussbaum, M. (2000). Voices of the poor, can anyone hear us. Oxford, England:
Oxford University Press.
Pallant, J. (2007). SPSS Survival manual: A step by step guide to data analysis using
SPSS Window. New York, NY: McGraw Hill/Open University Press.
Peat, J. (2001). Health science research. A handbook of quantitative methods. Sydney,
Australia: Allen & Unwin.
Rank, M. (2004). One nation underprivileged: Why American poverty affects us all.
New York, NY: Oxford University Press.
Ravallion, M. (1998). Poverty line in theory and practice (LSMS Working Paper, 133).
Washington, DC: World Bank.
Saith, R. (2001). Capabilities: The concept and its operation. Retrieved from http://
www3.qeh.ox.ac.uk/pdf/qehwp/qehwps66.pdf.
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
Choosing Concepts and Measurements of Poverty 15
Schiller, R. B. (2008). The economics of poverty and discrimination. Upper Saddle
River, NJ: Pearson Prentice Hall.
Sen, A. (1985). Commodity capabilities. Amsterdam, the Netherlands: Elsevier.
Silver, H., & Miller, S. (2003). Social exclusion: The European approach to
social disadvantage. Retrieved from www.brown.edu/Departments/Sociology/
faculty/hsilver/documents/silver_and_miller-european_approach_to_social_
disadvantage.pdf
UNDP. (1997). Human development report 1997 . New York, NY: Oxford University
Press.
UNDP. (2001). Human development report 2001. New York, NY: Oxford University
Press.
U.S. Bureau of Census. (2004). Poverty threshold 2004. Retrieved from http://www.
census.gov/hhes/www/poverty/data/threshld/thresh04.html
U.S. Bureau of Census. (2005). Current population survey: Annual social and
economic supplement. Retrieved from www./pubdb3.census.gov/
U.S. Bureau of Census. (2007). 2004 American Community Survey: Selected profile.
Retrieved from http://factfinder.census.gov/servlet/DTGeoSearchByListServlet?
ds_name=ACS_2004_EST_G00_&_lang=en&_ts=276021053828
U.S. Bureau of Census. (2008). American Community Survey: Poverty 2007 and
2008. Retrieved from www.census.gov/prod/2009pubs/acsbr08-1.pdf
U.S. Department of Health and Human Services. (2007). The 2007 Health
and Human Services guidelines. Retrieved from http://aspe.hhs.gov/poverty/
07poverty.shtml.
U.S. Department of Health and Human Services (2009). The 2009 HHS poverty
guidelines: One version of the US federal poverty measure. Retrieved from
http://aspe.hhs.gov/poverty/09poverty.shtml.
U.S. Department of Labor. (2006). NLS web investigator. Retrieved from
http://www.bls.gov/nls/
World Bank. (2002). Measuring poverty. Retrieved from www.web.worldbank.org/
WBSITE/EXTERNAL/TOPICS/EXTPOVERTY/EXTPA/0,contentMDK:20202198∼
menuPK:435055∼pagePK:148956∼piPK:216618∼theSitePK:430367,00.html
Downloaded by [Moses Kwadzo] at 12:56 10 July 2015
