Paper short abstract:

In this work we seek to show the process of creating a Multidimensional Poverty Index for the state of Minas Gerais, in Brazil; the advantages of this index compared to other approaches; and how this index was used by the government of this province in public policies in the area of combating poverty and promoting human development.

Paper long abstract:

In this work we seek, at first, to show how was the process of creating a Multidimensional Poverty Index (MPI) for the state of Minas Gerais, in Brazil; and how this index was used in public policies that aim to eradicate poverty and promote human development by the government of this province. In a second moment we will focus on the differences in poverty measurement results when we use the MPI and when we use income alone. Afterwards, we analyze how the phenomenon of poverty changed from these different perspectives in the state of Minas Gerais from the beginning of the pandemic until the end of 2023. We also show the challenges of creating a multidimensional poverty index from CADÚNICO, a database that works as a census of the most vulnerable population in Brazil.

Minas Gerais is the fourth largest state in Brazil, with more than 586 thousand square kilometers and a population of more than 21 million people.

The Minas Gerais Multidimensional Poverty Index (also called "Minas MPI") is an index composed of 12 indicators, covering 4 dimensions: Basic sanitation, Education, Standard of living and Work and Income. To build it we used Cadúnico. The Cadastro Único for Social Programs (CadÚnico) is a tool used in Brazil to identify and characterize low-income families. It is managed by the Federal Government and operates as a database containing socio-economic information of Brazilian families in situations of social vulnerability. The CadÚnico is used as a criterion for the selection of various social programs and benefits, such as the Bolsa Família. The cadúnico has its own poverty line and we use these lines to measure poverty through monetary approach and make a comparison with the multidimensional index built from this same database.

Firstly, we will talk about the challenges in creating this type of index using a database like CADÚNICO. We will also show what this index can be used in the context of public policies that aim to tackle poverty and promote human development.

In the second moment we will show the data generated from the two approaches to measuring poverty: the multidimensional approach and the monetary approach. To summarize, these data show an increase in poverty in the state of Minas Gerais during the Covid-19 pandemic. But when we use IPM, this increase is smaller. The two approaches have similarities when we look at the regional distribution of this increasing trend. The increase, both in the monetary and multidimensional approaches, occurs in regions that historically suffer less from the problem of poverty: the southern, southwestern and central-western regions of the state. We attribute this greater increase in these regions during this pandemic period to two factors: first, the fact that the "poverty ceiling" is higher in them than in other regions of the province; second, greater institutional learning from public facilities in other regions to deal with times of increasing poverty.

The data also show that the increase in multidimensional poverty occurs only in indicators related to the dimension of work and the dimension of education. In other indicators there is a reduction in the percentage of families in deprivation.This happens because there is a massive entry into the Cadúnico of families who already had access to housing, basic sanitation and education, but who lost income and work during the pandemic. This phenomenon is a symptom of a process of impoverishment of sectors of the middle class and lower middle class during the pandemic and the economic crisis generated by it.

These findings shows us some advantages and disadvantages of a multidimensional indicator. It allows, on the one hand, to capture people's poverty in a granular way. It is possible to differentiate in which dimensions families are deprived, which, in turn, allows the construction of public policies that are more appropriate and adherent to the specific situation of each family. On the other hand, a multidimensional indicator like the MPI tends to be less sensitive to changes in urban contexts or middle-class contexts. In this type of context, most or all families tend to have access to basic urban infrastructure, so that families that experience a worsening of income and/or unemployment, depending on the indicators and the cutoff line used to define those who are deprived or not, may not be identified as vulnerable/poor families.