Advances in sub National measurement of the Human Development Index: The case of Mexico
This paper examines the main adjustments made to the HDI (human development index) in the Mexican National Human Reports in terms of analyzing development levels for subgroups of population either by age, ethnics, sex and income made at the household and at individual level. The HDI proposes three main indicators, human development research bases on: the possibility of a long and healthy life, the acquisition of a good knowledge and the opportunity to get a respectable standard of living. All these data are reported to inform the global Human Development Report which is published annually. Sub national analysis of the HDI inMexicostarts through regions which are composed of groups of 31 states and theFederal District. However, in this section, a basic aspect of human development is missing, that is the inequality between persons or groups and its achievements. Other surveys made to the HDI include oil revenues inMexicowhich are in the hands of the Federal Government and so in the public purse which redistributes them to the states according to budget allocation formulas. The adjustment consists on deducing the amount of oil revenues passing from oil producing states to the Federal Government and then adding the amount of these resources allocated to all the states.
As for the World Bank, globalization and the industrialization have increased the prices of oil, metals, and minerals since 2002 and as a result oil and metal importing economies have seen price increases. Another matter analyzed by HDI is the case of migration .When migration occurs, from one state to another, the states of origin and destination change because of different forces in motion. First, the travelling of human beings from one place to another modifies the cultural characteristics of migration people as well as new market conditions occur due to the new supply and demand of labour and goods of migrants. The analysis which is carried out, determines that the effects of migration of the migrants-less comparison group on the HDI, are negative for many states of the country.
However, this paper does not mean that the existence of migration is harmful for the migrants, in fact the average HDI decreases if a migrant with higher than average education index departs to a state where he/she has a lower than average index. So, the next natural step in this analysis, would be the connection between migration movements with the HDI changes.
InMexico, as in many other countries, there are some difficulties to calculate the municipal HDI .Even if different methodologies have been used for this purpose, none has been reliable. The only available one used to find information at this level is census data. But census data is particularly unreliable to get an index of available resources, so the solution that has been found to find out an estimation of per capita average household income is combining census and income surveys. The results of this analysis have been not too encouraging since a great inequality emerged illustrated by the fact that if municipalities were classified as countries, we would have some municipalities similar toItaly, and others similar toMalawi.
Since the HDI can be seen as the sum of three components (health, education and income indices), to identify which source has more importance on overall inequality of municipalities, it is possible to apply inequality decomposition techniques. One of these, uses the coefficient of variation which allows obtaining that the percentage of inequality derives (32.9%) from education, while the (30%) derives from health. In general these calculations are used to estimate those public actions which are necessary in order to increase the development. In 2010 the program against poverty in 28 municipalities was guided using the HDI. This same estimation might be applied in other countries likeEcuador,South Africa,Brasil,Panama,Albania,Bolivia,Indonesia,MoroccoandVietnam.
From the previous analysis, it is deduced that the HDI is a useful measurement that influences public policies however, it has many limits because it does not explains all the variables that influence the human development. So, in order to search for a complete measure of human development, it is possible to add other variables to see how they influence the HDI. For instance, the 2004 National Report considered essential particularly those data related to public security. A new public security dimension has been introduced and added to the HDI with the same weight as the health, knowledge and resources dimensions. With this dimension, many countries lose their place in the table. A very similar analysis has been carried out by PNUD in 2009 introducing the absence of violence against women as a new dimension. The results have been the same.
In conclusion, the HDI overlooks important dimensions of human development but when introducing new variables through sensitive analysis, the results have changed.
However the HDI fails to describe the inequality between individuals and dimensions and so it is not really representative of those countries with a high inequality likeMexico. Another problematic aspect of the HDI is its aggregation method that combines the data into an overall index. Anand and Sen and Hicks have proposed useful distribution-sensitive measures of human development trying to get through these limits. They have formulated axioms for a more complete index. These axioms are.1) Symetry in dimension 2)Symetry in population 3)Replication invariance 4)Monotonicity 5)Homonogeneity 6)Normalization 7)Continuity 8)Subgroup consistency 9)Transfer principle.
The innovation of this index called H(e) is the use of a distribution-sensitive general mean to summarize the dimension-specific level of human development.(e) can be interpreted as an “inequality aversion parameter”. For instance if two groups have the same HDI but one has a more unequal distribution, this will involve a lower H(e) but the inequality aversion parameter is bigger. This kind of analysis has been used to value seven Latin America countries (Nicaragua,Paraguay,Brazil,Dominican Republic,Uruguay,Argentina, andChile).
Different approaches have been used attempting to go beyond limits of the HDI.
One proposal is the creation of an inequality sensitive development index based on the concept of generalized means.
Another is used in the Well-O-Meter of the American Human Development Project. It consists of an index by asking 25 questions. The problem is how to weight each question and how all the data is aggregated.
Another shortcoming in the approach that uses groups to identify the classic variables of the HDI, is that certain individuals do not clearly identify their human development. When disaggregation occurs at the household and individual level, as in the Well-O-Meter analysis, considerations should be taken by summing up the basis of state and municipal disaggregation of the HDI. Usually, national and state data allow calculating HDI according UNDP methodology but for deeper levels of disaggregation. However, this analysis is not complete and does not carefully identify the income. This paper shows some of these methodological decisions at different levels of aggregation inMexico.
Life expectancy index
The life expectancy index computation, changes due to the different levels of aggregation. In fact as described in the paper at national and state level, it considers life expectancy at birth, while at household level it is necessary to introduce life tables for age and gender. This allows international comparisons. Besides, this paper describes sources of information about life tables in Mexico. The most recent and complete estimations are CONAPO’s which is charged of computing life expectancy and infant mortality rates. But the rates, do not consider the different income levels and in order to overcome this problem a stage linear regression model is carried out to sensitize life expectancy.
It is suitable to determine a maximum and a minimum life expectancy for each age and gender group as UNDP made for life expectancy at birth. All WHO’s estimates include these data. It is necessary to conciliate UNDP analysis with WHO analysis in order to build an adjusted index which can be comparable at international level. The purpose of this analysis is to compute the life expectancy index for individual influenced by the adjusted income for age, gender and state.
Following UNDP, the variables that influence knowledge are literacy and the years of schooling adjusted for each individual’s age. However, it is necessary to adjust these values. Another methodology for calculating the education index considers two indices, one for adult literacy and another for combined gross enrolment. It assigns two-thirds weight to the first and one- third weight to the second to create the education index. To extend this computation at household level, it should be considered:
The final objective is to calculate the education index of each household member.
Available resources index
Traditionally GDP index is calculated using Gross Domestic Product at purchasing power parity. The concept of income at a household level forMexico, is calculated by TCPM, a committee composed by scholars and created by the federal government to define official poverty measurement.
Monetary and non-monetary resources are being used to calculate the total current income. Subsequently the income is adjusted to national accounts and to the current income to obtain a comparable measure at international level. Once the dimension indices have been calculated at household level, it is straightforward to calculate the HDI index. It is a simple average of the three dimension indices:
HDIh=⅓(Life expectancy indexh ) +⅓(Education indexh )+⅓(GDP indexh )
The main themes of the paper the adjustments made to the HDI in the Mexican National Human Development Report and their use; the calculation of the HDI carried out to the individual and household level. These adjustments include the redistribution of GDP and imputing per capita average household income to all municipalities combining census and income surveys to analyze regional inequality. Another adjustment have been made introducing new dimensions as local crime and the absence of violence, which allowed to obtain useful conclusions for public policy in Mexico. A key contribution to the HDI analysis was given by approach based on the concept of generalized means, and by a new way to disaggregate the HDI at the household and individual level from income survey data. The proposal of a rigorous inequality sensitive HDI and the household and individual calculation of HDI is important innovation that contribute to improve the analysis of human development. Despite the efforts for the adjustment of this index, as the case of Mexico, the HDI continues to receive much criticism, because first of all the HDI is an agglomerative index that trends to not reflect the objectivity of the data, but also because of the choice about dimensions, indicators and weight assigned to the selected size.