Built-in statistical technology with user-friendly interface
Customizing programs to easily meet local needs
Washington, January 5, 2012 — As the World Bank goes from open data to open knowledge, a free software program created by Bank researchers is offering staff and policy makers around the world a helping hand in evidence-based decision making.
The ADePT software allows users, with just a few mouse clicks, to generate reports using data from sources such as household surveys — all presented in print-ready, standardized tables and charts. It can also be used to simulate the impact of economic shocks, farm subsidies, cash transfers, and other policy instruments on poverty, inequality and labor. The software not only automates the analysis, but helps cut down on human error and introduces new methods of economic analysis to researchers worldwide.
The free software is especially valuable for analysts in developing countries, where expensive statistical software tools and training are often hard to come by. With built-in modern statistical technology and a user-friendly interface, ADePT empowers policy practitioners — including those with limited programming skills — to conduct sophisticated economic analysis. It also serves as a platform for researchers to share findings.
“Our goal is to develop modern, user-friendly tools to bridge the gap between the needs of busy policy makers and the vast ocean of data already out there,” says Michael “Misha” Lokshin, the main creator of the program and an economist at the World Bank’s research department, the Development Research Group. “When local planners can easily calculate statistics based on survey data, it will ultimately help governments make informed decisions and design better policies.”
The initiative comes as developing countries see higher demand for transparency and data sharing. The World Bank, which already offers its data free to the world, is focusing on sharing its expert knowledge as well. The Development Research Group offers training courses, how-to guides, databases, and other tools for knowledge transfer.
But rarely has a knowledge-sharing project gone as far as commercial-grade ADePT. So far, ADePT has attracted about 10,000 users, and the number is growing as new modules, functionality and features are introduced. It now includes applications specifically designed for the analysis of poverty, inequality, social protection, labor, health, education, and gender. It is also being adapted for specialized use by various countries, economic sectors and institutions, such as the United Nation’s Food and Agriculture Organization and International Labor Organization.
In the last four years, the ADePT team, led by Lokshin, Zurab Sajaia and Sergei Radyakin, has trained analysts in government, academia and think tanks from about 30 countries, including Indonesia, Russia, Mexico, and South Africa, on how to use the program.
They are working with other economists to produce a series of manuals and step-by-step guides, Streamlined Analysis with ADePT, to accompany the sector-specific applications. The first two volumes, on education and health, have already been published. In addition, the ADePT team has developed customized interface in 10 languages to better meet the needs of policy makers in the World Bank’s client countries.
ADePT began as a small-scale project. As an economist specializing in applied economic research, Lokshin saw firsthand the need to computerize the process of producing standard Bank reports. But it was a gamble to develop such software: Many were worried that an easier process would lead to careless production of reports with bad data. Lokshin and Sajaia would also need to build the software from scratch, work directly with potential client governments and manage projects, marketing and customer service.
With a shoestring budget, they completed the first version of ADePT in fall 2007. For months, they conducted seminars and invited people to try out the new software. Slowly, they won over skeptics. In March 2010, they negotiated with StataCorp, the statistical software company, to purchase its computational engine, Numerics by Stata. They then embedded it into ADePT so everyone can use it for free and not have to pay a $1,500 licensing fee, a hefty sum for individual users and those in developing countries. That paved the way for the mass distribution of the software.
“ADePT is a great example of what I call ‘wholesaling research,’ whereby good researchers turn some of their skills and energy to the task of producing the data and software needed by other researchers,” says Martin Ravallion, director of the research department. “Thus we can help build the capacity for evidence-based policy making in developing countries by reducing the cost of accessing cutting-edge methods.”