Distribution of EU Aid in Mexico: A Preliminary Look

The Fellowship ended August 8th, with several maps half made and a report half-finished. We simply ran out of time as Fellows, although with the trainings that we conducted over the summer for both geocoding and geospatial analysis, ObservaCoop should be equipped to create basic maps for the report comparing the distribution of European Commission aid projects in Mexico with indicators such as poverty, violence, education, and more. Although I was not necessarily permitted to publish those maps, I did have access to the geocoded data and consequently constructed an elementary model of EU aid distribution at the state level using tools from my econometrics course that I took spring of this year.

In the weeks after the end of the geocoding portion, I spent a good deal of time collecting data at both the state and municipal level for variables that included basic indicators like poverty, primary education, and HIV/AIDS mortality rate but also for more ambiguous indicators such as “culture,” which was represented by the number of libraries and books that each county had. When we left, each Fellow had an impressive set of indicator data for the states of Mexico as well as municipal data for Oaxaca, Chiapas, and Tamaulipas, which had a large number of project locations and therefore merited sub-state analysis. After leaving Mexico, I took a look at my data and decided to investigate the distribution of EU aid projects with relation to the Millennium Development Goals indicators, many of which we had found data for in our dataset.

Of course, I wasn’t able to find data on all of the MDG indicators at the state level for Mexico. Going through the current literature for aid allocation and MDG indicators at the national level (sub-national analyses are rare since data is sparse), I decided to comb the Instituto Nacional de Estadística y Geografía (INEGI), Mexico’s national statistical agency, for indicators for five of the eight MDGs, as illustrated below.

GOAL

VARIABLES (per State)

Goal 1: Eradicate extreme poverty and hunger
  • Gross domestic product
  • Number of those living in poverty
  • Percent of population lacking access to food
  • Population
  • Number of people over 14 unemployed  and looking for jobs
  • Number of people over 14 unemployed and not looking for jobs
Goal 2: Achieve universal primary education
  • Number of students enrolled in primary school
  • Number of people who have finished all years of primary education
  • Number of people who are illiterate
Goal 4:Reduce child mortality
  • Number of children less than one that died
Goal 5: Improve maternal health
  • Number of mothers who died
  • Number of teen births
Goal 6:Combat HIV/AIDs, malaria and other diseases
  • Number of victim that die from AIDS
  • Number of people that lack access to basic healthcare
  • Access to clean water

 

Using the general to specific method to isolate a set of the most influential determinants of European aid, I was able to identify maternal mortality, access to basic health services, and GDP as the most significant indicators. Of course, this doesn’t imply that the other indicators are not important in the EU’s consideration of aid; rather, these three variables have the greatest relative significance. If anything, they are highly correlated with several of the other indicators. For example, to reduce maternal mortality it is not necessary to provide health services to mothers but also to provide them with adequate nutrition, for example. The relationship between GDP and projects, which was curiously positive, could be explained by strong government institutions or infrastructure in those states, which would simultaneously result in higher GDP and a better environment for implementing aid projects.

This preliminary study is certainly not perfect and suffers from lack of data and more advanced methods for statistical analysis. However, the study itself is potentially significant because it uses subnational aid data to investigate allocation patterns, which in the past have been plagued with geopolitical factors between countries that overshadow the effects of recipient need on the distribution of aid. By investigating projects within a country, it is possible to hold constant those factors and see if aid from donors is more altruistic than they appear in past studies.

As the next chapter of my college education begins, I hope to learn the tools that would help me better investigate aid allocation. Furthermore, as more granular data from donors come forth in response to the demand created by institutions such as AidData and ObservaCoop, it may be possible to revisit this Mexico case in the future and refine my analysis.