Back to Programme

Comparing Polling-based API (Absolute Poverty Index) to the Human Development Index (HDI) in India

Yashwant Deshmukh (CVoter Foundation)

Keywords: Survey research and questionnaire design

Abstract

Based on 2016 findings from the Pilot ‘Survey of the Poor’ in India, this paper describes the development of an Absolute Poverty Index based on subjective and objective measures of well-being. The paper then does a state-by-state comparison of this index with the well-established Human Development Index (HDI), widely used by international development and aid organizations.

The discussion on Poor and Poverty has always been a skewed one. There could never be one universal definition of Poverty, leave alone the absolute poverty. Several attempts to do the same have only resulted in generalization of “Perceived Poverty” or perceived “Quality of Life” from the western parameters, which only compounded the problem further by sweeping the “Extreme Poverty” under the same definition, lock stock and barrel.

The ideas on developing various indices have only taken the discussion to next level as the age old parameters of mapping “Development” have been challenged by looking “Beyond GDP”. In such a scenario, it only seemed fit to pick up what all these indices wanted to map, some overlapping, and some exclusive. All those parameters, when put in one place were looking chalk and cheese. So in order to make sense of all these, in my model of Binary Composite Indicator; we grouped SOP questions in the following ten parameters; 5 being objective and 5 subjective.

Objective Indicators:
1. Food
2. Clothing
3. Shelter
4. Healthcare
5. Education

Subjective Indicators:
6. Safety
7. Self esteem
8. Gender equation
9. Equal opportunity
10. Optimism

All these indicators have been directly or indirectly covered (more or less) in the currently proposed SOP questionnaire. Though the construct of the questionnaire in non-linear; in other words most questions have exclusive atypical design from each other. However; I believe we can still derive enough data to construct the model of Binary Composite Indicator.

Here is my attempt at calculating and configuring an API based on different questions asked in the SOP Pilot survey. As I have tried to construct 5 Objective Indicators : Food, Clothing, Shelter, Healthcare and Education on a binary scale (1= Have & 0= Have nots) followed by constructing 5 Subjective Indicators : Safety, Dignity, Gender Equality, Equal Opportunity and Optimism on a binary scale (1= Have & 0= Have nots) .

By adding these two to make a Composite Indicator; we finally get the API on a 0 to 10 scale. That scale could be used not just to rank any Geographic of Demographic Unit; but also, to study the relative sensitivity of the two scales with each other. This goes remarkably in line with the HDI ranking of different states in India; with few contrasting cases where Industrial growth or GDP growth has not worked in sync with the Poverty reduction. In fact is only has increased the gap between rich and the poor and that is and that is extremely well reflected in our API. This paper attempts to explore the same with the primary SOP pilot field data and the available secondary indices.