A framework to calculate Multidimensional Poverty Index (MPI) by using Alkire-Foster method
Given N individuals, each person has D indicators of deprivation, the package compute MPI value to represent the degree of poverty in a population.
The inputs are 1) an N by D matrix, which has the element (i, j) represents whether an individual i is deprived in an indicator j (1 is deprived and 0 is not deprived) 2) the deprivation threshold.
The main output is the MPI value, which has the range between zero and one. MPI value is approaching one if almost all people are deprived in all indicators, and it is approaching zero if almost no people are deprived in any indicator.
You can install the newest version from Github.
devtools::install_github('9POINTEIGHT/MPI')First we will use simulation poverty data from build-in package. Which contains 30 rows of individuals, 16 columns of deprivatied dimensions (1 is deprived and 0 is not deprived), and simulated forth-level administrative division of France.
MPI::examplePovertydf
We use the following function to compute MPI.
out_seq <- AF_Seq(df = examplePovertydf, g = "Region", k = 3)Input will be… * df A poverty data frame *
g A column name that will be used to divide data into
groups. When the value is NULL, the entire data is not separated into
groups.(default as NULL) * w An indicator weight vectors
(default as 1) * k A poverty cut-off. If an aggregate value
of indicators of a specific person is above or equal the value of k,
then this person is considered to be a poor.(default as 1)
Output will be list of lists separated into group, and
each list contains… * groupname A Grouped value from column
input g * total Number of population in each
group * poors Number of deprived people in each group *
H Head count Ratio, the proportion of the population that
is multidimensionally deprived calculated by dividing the number of poor
people with the total number of people. * A Average
deprivation share among poor people, by aggregating the proportion of
total deprivations each person and dividing by the total number of poor
people. * M0 Multidimensional Poverty Index, calculated by
H times A.
[[1]]
[[1]]$groupname
[1] "Bastia"
[[1]]$total
[1] 2
[[1]]$poors
[1] 2
[[1]]$H
[1] 1
[[1]]$A
[1] 0.4090909
[[1]]$M0
[1] 0.4090909DimentionalContribution
indnames The poverty indicatorsdiCont Dimensional contributions denotes the magnitude
of each indicator impacts on MPI.UncensoredHCount Uncensored head count of indicator
denotes the population that are deprived in that indicator.UncensoredHRatio Uncensored head count ratio of
indicator denotes the proportion of the population deprived in that
indicator.CensoredHCount Censored head count of indicator denotes
the population that are multidimensionally poor and deprived in that
indicator at the same time.CensoredHRatio Censored head count ratio of indicator
denotes the proportion that is multidimensionally poor and deprived in
that indicator at the same time.
pov_df poverty data frame
Cvector is a vector of total values of deprived
indicators adjusted by weight of indicators. Each element in Cvector
represents a total value of each individual.IsPoverty is a binary variable (1 and 0). 1 indicates
that a person does not meet the threshold (poor person) and 0 indicates
the opposite.Intensity , The intensity of a deprived indication
among impoverished people is computed by dividing the number of deprived
indicators by the total number of indicators.
Alkire S., Chatterjee, M., Conconi, A., Seth, S. and Ana Vaz (2014) Global Multidimensional Poverty Index 2014. OPHI Briefing 21, Oxford: University of Oxford.
Please, visit Global Multidimensional Poverty Index 2014