MPI : Multidimensional Poverty Index (Alkire-Foster Method)

The mpi package provides tools to compute the Multidimensional Poverty Index (MPI) using the Alkire-Foster methodology.

It allows users to input binary deprivation indicators, assign weights, and specify a global poverty cutoff (k). The package returns:

• A data frame with individual deprivation scores and poverty status
• A summary table with key statistics: Incidence (H), Intensity (A), and MPI
• An interactive plotly barplot of poverty status

Installation

# Install from GitHub
# remotes::install_github("julienParfait/MPI")

Example

library(multidimpoverty)

data <- data.frame(
  edu = c(1, 0, 1, 1, 0),
  health = c(0, 1, 1, 1, 0),
  water = c(1, 0, 1, 1, 1)
)

weights <- c(0.4, 0.3, 0.3)
k <- 0.33

res <- mpi(data, weights, k)

head(res$data)

##   edu health water DeprivationScore IsPoor
## 1   1      0     1              0.7      1
## 2   0      1     0              0.3      0
## 3   1      1     1              1.0      1
## 4   1      1     1              1.0      1
## 5   0      0     1              0.3      0

res$summary

##                                 Metric Value
## 1                  Headcount Ratio (H)  0.60
## 2                        Intensity (A)  0.90
## 3 Multidimensional Poverty Index (MPI)  0.54

#res$plot

Output Meaning

• H (Headcount ratio): Proportion of people who are multidimensionally poor

• A (Intensity): Average share of deprivations among the poor

• MPI: The product of H and A