In 1997, MI developed a T21 model for Malawi to help its government translate the Vision 2020 goals into measurable objectives through national stakeholder consultations and analysis of scenarios. The outcome was a new national development strategy, Reaching the Vision that sets out the path to attaining the national vision.
For past projections, the model performs well for total population with an average deviation smaller than 5% from UNPOP and WDI values. For real GDP, the dips in 1992 and 1994 are caused by severe droughts that occurred in Malawi and the model didn’t predict.
For 1998-2006, the total population still stays within 5% and for GDP, the model is able to reproduce their medium to longer term quite well, but underestimates growth after 2001.
In 1998, under a contract with ANPA, Italy's National Agency for the Protection of the Environment, and with collaboration from ENEA, the Italian Department for New Technology, Energy, and the Environment, MI customized the T21 template model to Italy and began an exploration of how best Italy could achieve its various international environmental commitments. The goal was to find scenarios under which Italy could achieve its commitments without doing serious damage to its economy.
The Italy model performs very well against actual data. For past projections, total population has an average deviation of 2% and Real GDP is about exactly the same as WDI.
For 1999-2006, total population remains within 2% until 2003 and after that is still 4-5%. The Real GDP remains about 4% throughout, becoming better over time.
In 2002, General Motors supported the development of T21-China for highlighting China’s growing energy and food demand.
The T21-China model performs very well when compared against the actual data. For past projections, the values are within 2% for total population and Real GDP varies between 2-4%.
For 2003-2006, total population also has an average deviation of only 2% and for Real GDP the model underestimates the economic growth driven by government actions not accounted for when building the model.
In 2002, MI created the T21-Thailand model to look at population, reproductive health, and HIV/AIDS.
In terms of past and future projection, total population has an average deviation of only about 2%. The simulation of the HIV Adult Prevalence Rate has a difference of only about 0.2% both past and future, but captures the longer-term trend pretty well, especially considering the inconsistencies in UNDP and UNAIDS data.
T21-Papua New Guinea, 2002
In 2002, Conservation International and MI collaborated on pursuing a more cooperative approach to address the concerns of various interest groups represented in Papua’s environmental and economic resources to create T21-Papua: A new approach to integrating development planning with biodiversity conservation.
For past projection, both population and Real GDP fall within 5% of actual data. For 2003-2006, population remains within 5% but for Real GDP a spike occurs in 2004.
As for total forest land, the model performs very well both past and future in spite of a change in classification methods at FAO. As a consequence, the simulation is consistently about 5% lower than Total Forest and Wooded land, but 5% higher than Total Forest Land, perfectly matching the long term trend.
In 2002, MI and the Government of Bhutan (GoB) collaborated to create a T21 model. In 2004, as part of the Netherlands Climate Change Studies Assistance Program, the GoB decided to use T21 to investigate impacts of climate change on Bhutan.
For past projection, the model accurately simulates the trend of total population and GDP falls within 5%. For 2003-2006, total population continues to represent the trend in WDI and UNPOP data and for GDP, the rapid increase in 2006 is due to the completion of a major hydroelectric plant.
T21-Cape Verde, 2003
In 2003, Senior Cape Verdean government officials identified T21 as an excellent tool to assist in undertaking integrated strategic planning, involving diverse stakeholders in the planning process, and monitoring performance against agreed goals. MI developed T21 Cape Verde specifically to support the Poverty Reduction Strategy Paper (PRSP) process.
In terms of past projection, for total population there is an average deviation of 3% with UNPOP data and for Real GDP, it is also only 3% from IMF and WDI data as shown.
For 2004-2006, the total population is now within 5% of actual data and Real GDP remains within 3% of WDI data.
In 2003, MI created T21-Ghana (Assessing best options for meeting the Millennium Development Goals in Ghana) in order to assess the impact of MDG-related interventions on the national economic and social development, and the synergies (or lack thereof) among them.
In terms of past projection, for total population there is an average deviation of only 2% from UNPOP and WDI data and for Real GDP, the projection falls to within 3%.
For 2004-2006, population is within a maximum of 3% and GDP a maximum of 4%.
In 2003, MI worked with Mozambique’s government ministries and civil society groups to build their capacity to use T21, and use it as a framework for broad dialogue on policy issues, thus increasing broad participation in national planning.
For past projection, the total population falls to within 4% of UNPOP and WDI measures and the Real GDP varies a bit but is mostly within 4% of actual data.
For 2004-2006, population has a deviation of between 4-5% and GDP falls consistently within 2-3% of actual data.
In 2003, under The Carter Center’s Development and Cooperation Initiative (DACI) with the Government of Mali, MI used T21 to support the preparation of Mali's poverty reduction strategy paper for the World Bank (PRSP).
In terms of past projection, for population there is an average deviation no more than 3% for both WDI and UN data and for Real GDP there is variation, though it follows a similar trend and is usually within 5% of WDI data.
For 2004-2006, population has 4% average deviation while GDP is within 3% of actual data.
The second version of the model for the United States, which focused on the economic sector and was created and featured on C-SPAN, overall performs very well on the major indicators.
For past projections, the total population falls to within 3% of UNPOP and WDI, total real trust funds have a average deviation of about 4%, and Real GDP has an average deviation of just 3%.
For 2005-2006, population remains within 3%, total real trust funds goes up to around 5%, and a spike for Real GDP occurs in 2003 due to the economic and monetary policies of the second Bush administration that have stimulated the financial sector and housing market more than expected.
T21-St. Lucia, 2004
T21-St. Lucia is a simplified T21 model that was developed to support a training program on integrated development planning conducted in the country in 2004.
For past projects, total population falls to within 1-2% of WDI data and Real GDP is within 2-3% of WDI data, while keeping in line with fluctuations that can occur but are difficult to account for.
For 2005-2006, the population has an average deviation of 3% from the WDI data and the Real GDP is within 4% of WDI data.