• Subscribe

Feature - Computing a way out of poverty

Feature - Computing a way out of poverty

In northwest Bangladesh, 72% of the population live below the poverty line, according to the Asian Development Bank. Photo here and on front page courtesy of ADB and Eric Sales, from an ADB Photo Essay entitled New Hope for Bangladesh's Extreme Poor.

In the Philippines, the Asian Development Bank and grid specialists are getting together to figure out best investment strategies for Asia's poorest regions.

One of the outcomes of the EUAsiaGrid Project, funded by the European Commission under Framework Program 7, has been to spark new grid-based collaborations in Southeast Asia.

One example is the result of a meeting that occurred last October, when representatives from EUAsiaGrid partners went to the Asian Development Bank (ADB) Headquarters in Manila, The Phillipines, to explore ways to model the impact of poverty alleviation investments that ADB makes.

In the past, investments by development banks such as the ADB and the World Bank have been guided by broad macroeconomic theories, with often controversial results.

But given detailed enough models, argues Hans-Peter Brunner, a senior economist at ADB, it should be possible to simulate microeconomic behavior down to the scale of individuals buying and selling in marketplaces. These are called "scaled agent-based models," because the computer program simulates the relevant economic interactions within a population of economic 'agents' representative of a given economic region.

A key advantage of such agent-based models is that they can reproduce the evolutionary nature of the economics of a given geographical region. In particular, the model has an explicit internal representation of space which can be related to the actual geography of the region. Hans-Peter Brunner and Peter Allen have pioneered development policy experiments with the help of such evolutionary trade models.

An ADB-assisted project is providing extremely poor households, which normally rely on women begging or doing hard labor for income, with land and livestock to earn income, for example from milk production. Photo courtesy of ADB and Eric Sales.

A realistic picture

Run such models long enough, under a wide enough set of conditions, and a statistically realistic picture of the long-term impact of a set of investments ought to emerge.

In particular, if the impacts of previous investments have been monitored, it should be possible to run the models and tweak them until they reproduce what actually happened in the past. This in turn gives modelers some confidence in the ability of their models to evaluate the future investment options.

There is one hitch with evolutionary trade models: The amount of simulations - and the large number of agents needed for each run - makes this a massive computing challenge. However, it is one that is ideally suited to cluster-based grid computing.

That is precisely why the grid experts were invited to the ADB meeting. Rafael Saldaña, a coordinator of the EUAsiaGrid project based at Ateneo de Manila University, now has a team collaborating with the ADB experts. The Washington-based Brookings Institution currently has another team assessing the mathematical and computational aspects of the evolutionary trade model.

Initial modeling will focus on Eastern India, Bangladesh, Bhutan and Nepal, some of the poorest and most populous regions in the world. But Saldaña notes that in the future, the techniques and insights gained from this sort of study should be applicable also to countries like the Philippines and other regions of Southeast Asia, which will surely create a lot of demand for the grid computing resources that have been set up by the EUAsiaGrid partners.

-Francois Grey reporting for EUAsiaGrid

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2022 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.


We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.