Using model to identify and meet potential challenges in regional
development: The ECOWAS T21 case
Weishuang Qu’, Herve Lohoues”, J ed Shilling', Matteo Pedercini!
‘Millennium Institute
1634 Eye Street, N.W.
Suite 300
Washington, DC 20006, USA
Telephone: 202-383-6200
Fax: 202-383-6209
Email: wq@ millennium-institute.org
"ECOWAS Commission
101, Y akubu Gowon Crescent,
P.M.B 401 — Abuja, Nigeria
Tel: +234 813 157 9061
Email: hlohoues@ ecowas.int
ABSTRACT
The ECOWAS region (Economic Community of West African States) has big potential, but it
faces major challenges in its development. To help support the decision making of regional and
national leaders and bring a wide variety of stakeholders from all member states into policy
debates, the ECOWAS T21 model was developed. The initial focus of the model was to test the
consequences of regional integration of 1) free movement of people and commodities; 2)
integrated energy, transport, and telecommunication infrastructures; and 3) creating a monetary
union. When building and calibrating the model, another challenge was identified: fast
population growth would make it difficult to improve the well being of the people in the region,
even with successful implementation of regional integration. As a result, a family planning
scenario was added. Results from the model show that a combination of regional integration and
family planning policies generates the best results: smaller population, longer life expectancy,
higher GDP, much higher per capita GDP, higher total government revenues, a lower poverty
rate, a lower unemployment rate, more forest land, and higher per capita cereal production.
However, any good policy could have its costs, such as higher oil demand and lower oil exports
in this case.
Key words: ECOWAS region, development planning, T21 model, scenario analysis, regional
integration, family planning
Ts Introduction
Covering a total area of 5,112,903 square kilometers with a total population of 300 million, the
ECOWAS region (Economic Community of West African States) is the most populous regional
economic community (REC) in Africa, comprising about 35 percent of sub-Sahara Africa’s
population (ECOWAS Vision Document 2011). The countries in the region include Benin,
Burkina Faso, Cape Verde, Cote d’Ivoire, The Gambia, Ghana, Guinea, Guinea Bissau, Liberia,
Mali, Niger, Nigeria, Senegal, Sierra Leone, and Togo.
ECOWAS is a highly complex region facing many possibilities and challenges. The regional
GDP is estimated at 157 billion in US$2001 in 2010 (ECOWAS statistical data office 2011), with
per capita GDP at US$523 (all currency units in this paper are based on constant 2001 values). The
majority of ECOWAS countries are classified as Least Developed Countries, and about 60% of
population lives in poverty (under US$1.25 per person per day). Continuing fast population growth is
expected in the coming years, with a total fertility rate of about 5.
Yet, the fundamentals for a better life abound. The ECOWAS countries are endowed with
considerable mineral, land, water, and human resources. There are gold, diamonds, uranium,
crude oil, and iron ore; and numerous waterways run across the region. ECOWAS States also
produce primary agricultural commodities, a significant amount of which is sold in international
markets.
The ECOWAS leaders and policy makers face major policy challenges. The first of these relates
to how the massive population will be fed, provided for, and educated; and how their human
capital will be utilized to achieve sustainable development of the region. The second relates to a
broader goal of how to articulate more coherent economic, social, and environmental policies
that foster more sustainable development and improved living standards in all the countries and
the region as a whole.
The mission of ECOWAS is to deal with these challenges, and it is believed that the most
important is to promote economic integration in all fields of economic activity. In 2007, the
Authority of the Heads of State and Government formulated a vision for the region, Vision 2020,
which aims at transforming the current “ECOWAS of States” into an “ECOWAS of People”, to
achieve a region without borders to derive maximum benefits from globalization. To bring
regional integration (RI) to fruition, it is essential to strengthen the decision making capacity of
the ECOWAS Commission with an advanced analytical tool that adequately captures the
socioeconomic dynamics of the region and supports looking into alternative future scenarios
resulting from policies focused more on their goals than simply continuing business as usual.
Such a tool would also help bring a wide variety of stakeholders, both State and non-State actors,
from all member states into policy debates, which will in tum enlist real support for more
mutually beneficial policy decisions and actions envisioned by ECOWAS.
The ECOWAS Community Development Program (CDP) team decided to choose the Threshold
21 (T21) Model as the policy tool to formulate a coherent program of actions for RI. The experts
in the CDP team worked with the Millennium Institute to identify the four pillars to start
modelling for RI. They are:
1. Free movements of people, goods and services, and capital among member states
2. Governance, peace, and security
3. Energy and infrastructure
4. Finance and monetary integration
The T21 model (Qu 2011) integrates a broad range of sectors in economic, environmental, and
social areas; and its transparent structure and user-interface enables all stakeholders to engage in
constructive dialogue about policy options, which helps achieve consensus. The model has been
applied in more than 20 countries and regions, and it has benefited from lessons leamed along
the line. In a few cases, T21 models were used to identify challenges the countries have to face in
2
the future (Qu 2005). In addition to building an effective country or regional model, a major
goal of every T21 project is to build the local team’s technical capacity to a level where they can
fully operate, update, and expand the model. They take over full responsibility for the model and
assure its continued use.
Il. The ECOWAS T21 Model
ECOWAS T21 is similar to most other T21 models except that 1) it is a regional development
model and all 15 countries are combined into a single entity; and 2) the main focus to address is
the possible consequences of RI.
Based on discussions with local experts, RI in ECOWAS could boost productivity, increase
employment, make better use of the potential hydro power, and reduce power transmission loss.
It could also have negative effects, such as lowering government tariff revenues, due to free
regional trade. However, it is possible to determine how much these losses in revenues would be
offset by higher revenues from increased employment and productivity, or other sources. A clear
advantage of a coherent dynamic model is its ability to identify these feedbacks and help find
ways to mitigate negative effects. Potential causal relationships from RI to these consequences
are shown in Figure 1.
Regional Integration Contents Consequences
Pilla]: Frbe iniowentionts Or pepper e MOemmE Reople
goods/services, and capitals
investment employment
free trade egional trade
total factor
Pillar II: Governance peace and security ——®peace and securi productivity factors Productivity
icati 0 Tevenues
; _-» communication net ment tax revenue
Pillar III: Energy and infrastructure
OS Ranspoiai fe investment efficiency
energy network transport efficiency higher hydro power
Pillar IV: Finance and monetary integration monetary union energy efficiency ——lower transmission loss
Figure 1: Causal diagram of RI pillars and their consequences
Starting from the first column on the left of Figure 1, the first policy variable “Pillar I: Free
movements of people, goods/services, and capitals” will have causal effects on the first two
variables in column two, “free movement of people” and “free trade”. Variables in column two
from left will affect variables in column three, such as “free movement of people” will affect
“regional trade” and “cost of productivity factors”. Finally, variables in column three will affect
the variables in the right column under the red title “Consequences”, which include
“employment”, “total factor productivity”, “tax revenues”, hydro power, and transmission loss.
This diagram shows how the RI policy measures will have causal impacts on these consequences.
The policy variables and the consequences in Figure 1 are built in the ECOWAS T21 model,
whose condensed structure is displayed in Figure 2.
Government —~
revenues
expenditures
investments
-
Jand
agriculture land
forest land
settlement land
Households ~~
income
investment
Per Capita Income
Gini coefficient
Figure 2: Overview of ECOWAS T21 model
Production of agriculture, industry, and services is at the center of Figure 2. This sector affects 1)
government revenues and expenditures above; 2) household income and investment below; 3)
energy demand and supply to the left, 4) international trade to the far left, and 5) resource
demand of land and water (not shown). Each of these sectors will further affect more sectors, as
the arrows show in the figure. From these links, it is possible to see how the increase in industrial
production and household income and consumption may increase revenues to offset the losses
from trade tariffs. Many feedback loops are formed in the model, such as from production to
government and household, then to total investment, then to productivity and employment, and
finally back to production.
The main issue of the model, RI, is represented by the pink box in the far right. It can directly
affect 4 sectors: government, energy, productivity and employment, and total investment.
Indirectly through further linkages and feedback loops, it affects almost all other variables in the
model.
Figure 2 is an overview of the model’s structure. The actual model is much more detailed to
simulate the real world and incorporate the quantitative causal relations among the variables.
Figure 3 is the actual industry sector in the T21 ECOWAS. Industry production is modeled
using the Cobb-Douglas production function, with capital, labor and total factor productivity.
The variables enclosed by <> are computed from other sectors of the model, and two variables,
industry production and capital industry, are computed in this sector and used by other sectors.
Total factor productivity is influenced by the effects from the four pillars of RI, as shown in the
left of the figure, and by other variables of health (represented by life expectancy), education
(represented by adult literacy rate), and infrastructure (represented by the density of functioning
infrastructure).
O——— p> Carital Industry
industry gross depreciation industry —__ AVERAGE LIFE
cavestmont_—> ita formation Saw (CAPITAL INDUSTRY
mea INDUSTRY CAPITAL
>
ea INITIALCAPITAL ELASTICITY
INDUSTRY ~~
<POLICY ~~ relative capital industry
<pilar 2 productivity effect> IMPLEMENTION: = industy _ ee
full regional integration Integration total factor i
<illar3 productivity effect>—» roductivity effect ‘Productivity Effect” "productivity industy relative industry
employment — yyrriat'inDuSTRY
pillar productivity effect> PRODUCTION
relative average life effect of health On
INITIALAVERAGE expectancy ppocietenly inieety ndusty —_aniALinpustry
LIFE EXPECTANCY employment> —~EMPLOYMENT>
HLASTICITY OF efectofedwatonon
PRODUCTIVITY TO LIFE ef of rastructure
average life EXPECTANCY INDUSTRY Productivity industry density on productivity
expectancy> industry
relative average adult
_ eee literacy rate
<average adult. a” ELASTICITY OF PRODUCTIVITY TO
ELASTICITY OF INFRA STRUCTURE DENSITY
literacy rate> ini, AAR PRODUCTIVITY TO ——_<Telative infrastructure INDUSTRY
ADULT LITERACY RATE EDUCATION INDUSTRY density>
Figure 3: Industry sector of T21 ECOWAS
Ill. Historical Simulation and Analysis
The ECOWAS T21 model starts its simulation from the year of 1990. Using data, local expert
knowledge and literature review for the historical period of 1990 — 2010, equations and
parameters in the model were modified and calibrated, and results from the model were
compared to historical data. These adjustments improve the realism of the causal relations in the
model to assure it provides an accurate representation of the economic, social, and environmental
relations. This calibration helps identify where conventional relations need to be modified to
better present real ECOWAS relations and improve data collection. After the historical fit is
done for all the major indicators whose data are available, the model is then run forward to
simulate until 2030 — assuming no major changes in business as usual — to determine how the
numerous endogenous variables and indicators will progress. Figure 4 presents the comparison
of industry production between simulated results (blue, for the entire period of 1990 - 2030) and
actual historical data (red, 1990 — 2010 only).
industry production
100 B
1990 2000 2010 2020 2030
Time (Y ear)
industry production : history\base —————————————_ usd01/Y ear
industry production : history\ECOWAS-Data ——————— usd01/Y ear
Figure 4: Comparing model behavior with historical data
While working on the historical calibration and simulation of the model, critical economic, social,
and environmental characteristics of the region became more apparent, especially the following:
1. Real GDP grew quite fast from US$ 65.1 billion in 1990 to 141 billion in 2008, at an
average annual rate of 4.4%. But due to high population growth, average annual per
capita GDP growth was only 1.7%.
2. Population grew from 181 million in 1990 to 288 million in 2008, at a high average
annual rate of 2.6%. The total fertility rate was high, and declined slowly, from 6.55 in
1990 to 5.5 in 2008. Life expectancy increased, but also at a slow pace, from 48.6 years
(female) and 46.3 years (male) in 1990 to 52.2 (female) and 50.5 (male) in 2008. 43% of
the population in 2010 was below the age of 15, indicating continuing population growth
in the coming decades.
3. The poverty situation is severe. Measured in US$ 1.25 per person per day, in ppp
(purchasing power parity), about 60% of the population was living under that rate in 2008.
4, Although government expenditure on education was low, about 2% of GDP, gross
enrollment rate (GER) for primary education was good and improving. In 1990 GER
was 70% for girls and 90% for boys. In 2008 it increased to 85% for girls and 96% for
boys. Adult literacy rate in 2008 was about 50% for female and 70% for male.
5. Productivity (per worker output) growth has been very slow. In the 17 years of 1990 to
2007 (employment data was only available for 1990, 1991, and 2007), productivity in
industry and services only increased 8% and 18% respectively, well under 1% per year.
Observed GDP growth is thus supported more by the expansion of the workforce, rather
than by productivity growth.
6. Unemployment rate has been quite low, such as 8.3% in 2007 (based on data, labor
participation rate is 0.63, i.e., 63% of adults with ages 15 and over are or want to be
working). This indicates, even with both parents working, the family is very likely living
under poverty, as poverty rate is 60%. This could be the consequence of two factors: low
salary level, and large family size (since the total fertility rate is over 5 in 2010).
7. Deforestation continues: forest land decreased from 91.6 million hectares in 1990 to 75.0
million hectares in 2008. Agriculture land (arable and pasture land) increased from 212
6
IV.
million hectares in 1990 to 249 million hectares in 2008. It seems that for every hectare
increase in agriculture land, half a hectare of forest is lost. Agriculture expansion could
be the major cause of deforestation in the region.
Even with the expansion of agriculture land, the situation of hunger has not improved
much. Per capita cereal production increased from 161 kg/person in 1990 to 192
kg/person in 2008, still much lower than the world average of 352 kg/person in 2007
(FAO 2011). The major indicator for agricultural productivity, cereal yield, increased
from about 0.90 tons/hectare in 1990 to 1.27 tons/hectare, still very low compared to
world average of 3.38 tons/hectare (FAO 2011).
Oil production, which is one of the major drivers of regional economy, seems to have
peaked in 2005 at 977 million barrels.
Future Challenges of the Business as Usual (BAU) Scenario
With the BAU scenario, the following challenges were identified for the long term (2010 —
2030). The numeric values of the relevant indicators are presented in the scenario comparison
tables in Section VI.
1.
Population will continue to grow fast, partly due to high total fertility rate, and partly due
to a high proportion of population under 15 at present. Total population of the region
could reach 480 million in 2030, 60% more than the current number of 300 million.
Fast population growth will continue to slow per capita GDP growth. Asa result, 39% of
population could still be living under the poverty line in 2030.
Due to population pressure, agriculture land and settlement land will continue to grow,
taking away land from forest and all other types of land. Forest land could decrease to 50
million hectares in 2030, from over 70 million hectares at present.
Agriculture yield has been growing and is still quite low. Even if yield continues to grow
at about the same rate in the past, food production, measured on per capita basis, would
only grow marginally. This means that feeding the population would remain a big
challenge.
Educating the young would be challenging as well, as population growth will put lots of
pressure on government social services of education and health care. This pressure could
extend to other services provided by the government in areas like water, energy, and
transportation. Services in these areas are vital to the productivity growth.
Annual crude oil production would decrease from about 800 million barrels at present to
650 million barrels in 2030, while annual regional demand could increase from about 200
million barrels at present to over 300 million barrels in 2030. This means that the region
would have less oil to export (350 million barrels in 2030 compared to 600 million now.).
Primarily due to population growth, the governments of the region will have a challenging task
to feed the people, alleviate poverty, provide education and other social services, and promote
productivity in the competitive global market. The task would become especially tough when
natural resources of the region, represented by oil reserves and forest land, are being depleted.
V.
Assumptions of Regional Integration (RI) and Family Planning (FP) Scenarios
To help meet these challenges, three scenarios were developed with the T21 ECOWAS model.
They are: RI, FP, and RI&FP.
For the RI scenario, it is assumed that successful RI implementation will start in 2012 and the
four pillars will have the following consequences:
1. Pillar I: Free movements of people, goods, services, and capital will result in larger
markets, more competition, and better allocation of human and other resources. These
will increase productivity by 5% over the BAU Scenario. This increase will happen
gradually over a period of five years. All the productivity and foreign investment
increases in the following pillars will happen in the same pattern. Free regional trade will
reduce government revenues by 5% of inter regional import values, and the reduction will
happen as soon as the policy is implemented.
2. Pillar II: Governance, peace, and security mean that better governance and better use of
existing resources will increase productivity by 5% over the BAU Scenario; and it will
encourage more foreign investment, and as a result foreign investment will be 5% higher
than the BAU Scenario.
3. Pillar III: Energy and infrastructure of transport and telecom improvements will lead to
the construction and coordinated operations of regional integrated infrastructure networks
of energy, transport, and telecommunications, which will bring more reliability at a
reduced cost to the services from this infrastructure. As a result, productivity will
increase 10% over the BAU Scenario. These infrastructure networks will take five years
to build with an employment of 50,000 people. Power transmission loss will be gradually
reduced by 50% over the BAU Scenario in five years, and more hydropower will be
tapped (such as in Guinea) so that less oil will be used for electricity generation and more
oil will be available for export. Hydropower capacity will increase by 2% annually over
the BAU Scenario.
4. Pillar IV: Finance and monetary integration will bring more macro stability, and as a
result, both foreign investment and productivity will increase. It is assumed that they will
both be 5% higher than the BAU Scenario.
With the FP scenario, it is assumed that improved education and family planning programs will
help families make better decisions about the number of children they will have. As a result of
this program, it is assumed that the total fertility rate will decline from 5.38 in 2010 to 2.0 in
2030 in a straight line. In the BAU scenario, the total fertility rate in 2030 only declines to 4.12.
The scenario of RI&FP is simply the combination of all the assumptions in the RI and FP
Scenarios.
VI. Comparison of Scenarios
The model is built with the Vensim software, and all indicators from these four scenarios can be
examined either in graphical form or in tabular form. For instance, the indicator of per capita
real GDP, “real pe gdp”, from these scenarios is shown in Figure 5 in graphical form. The blue
line is from the BAU Scenario. The red line is from the RI Scenario. The green line is from the
FP Scenario, and the grey line is from the RI&FP Scenario.
|G. Obs isayoroe
2.000
1,500
1,000
500
9
poi 2011 2012 2013 201 S015 2016 2017 2018 s013 2000 021 2002 2023 d004 025 none 2077 9000 029 2030
‘Tine (Yeu)
realpe ap BAT ed Zearpertn)
rales RE ‘ect ea pets)
reaipe sep: FP sed Zeer person
realpo wp RIG? ved (Lear person)
(| | ew a5 Tle ]
Wise | Sidra | SATAECHWAC is | Biase nae mete | TA Greve o)a|a/<awe
Figure 5: Per capita GDP comparison of the four scenarios
A summary table comparing the results of major indicators for the years of 2030 is presented
below.
Table 1: Scenario comparison for 2030
Unit BAU Rl FP | RIGEP
Population:
Total population Millions 480 481 423 425
Average life expectancy years 56.31 58.07 57.03 58.76
Total fertility rate 4.12 4.05 2.00 2.00
Economy:
Real GDP Billion USS2001 391 589 395 596
Per capita GDP Uss2001/P 815 1,226 934 1,400
Government revenues Billion US$2001 70.3 93.7 71. 94.7
Social:
Poverty rate 39.09% | 21.29% | 32.95% 16.91%
Unemployment rate 7.55% 3.79% 6.86% 3.02%
Land:
Agriculture land Million Hectare 299 299 287 288
Forest land Million Hectare 50.1 50 55.6 55.4
Food and energy:
Per capita cereal production Kg/P. 248 313 280 353
Oil demand Million Barrels/Y 318 380 318 380
Oil export Million Barrels/Y 330 268 329 267
From the table we can see the differences among the scenarios in 2030. The RI&FP Scenario
generates the best results: smaller population, longer life expectancy, higher GDP, much higher
per capita GDP, higher total government revenues, a lower poverty rate, a lower unemployment
rate, more forest land, and higher per capita cereal production. However, any good policy could
have its costs, such as higher internal oil demand and lower oil export for the RI&FP Scenario in
the table.
People who are looking only from their specific positions could be concemed about short-term
losses from these alternative scenarios. For instance, due to free trade, government domestic
revenues from intra-regional import tariffs could decline. But in the long run, total government
domestic revenues will grow faster due to a larger tax base (GDP) and more imports from
outside the region, as Figure 6 shows.
jexooag Eo 29000
Comparison between Scenarios
ormesic reves
MODEL
1008
1B
2010 2011 2012 2017 204 201s ole 201? 2018 2019 2020 2021 2022 2002 2024 2025 2028 20z7 2028 2029 2030
‘Tame (Veet)
omestc revense BAT wa ter
omeirevenae: EE wes0iYear
— | Vow Tale ]
(ia | ones J rerecowastrsoant. | Nees clchanmade [fq Tr cEPAvean to @ e/a) «Gun
Figure 6: Comparison of government domestic revenues between BAU and RI
Beyond 2030, the differences among the scenarios could become even greater, as the total
population of the FP and RI&FP scenarios begin to stabilize, while it will continue to grow for
the BAU and RI scenarios with a strong inertia. The comparison of the population pyramids in
Figure 7 between the BAU and RI&FP scenarios in 2030 can show that in the next 20 years
beyond 2030, the RI&FP scenario will have a much smaller population of fertile women. With a
lower total fertility and a smaller fertile women population, the RI&FP will continue to generate
lower population growth. The higher per capita income achieved in the BAU and RI scenarios
lead to a very modest decline in the fertility rate, which is taken into account. But the differences
in total population and its growth are still quite significant between BAU and RI, and the FP
scenarios.
10
BAU
RFP
Population Pyramid for 2030
Female
age 80 and over
age 75-79
age 70-74
age 65-69
age 60-64
age 55-59
age 50-54
age 45-49
age 40-44
age 35-39
age 30-34
age 25-29
age 20-24
age 15-19
age 10-14
age 5-9
age 0-4
40M 30M 20M 10M 0 0 10M 20M 30M 40M
Figure 7: Comparison of population pyramids between BAU and RI&FP
VII. Summary and Further Work
It is shown in this paper that a good use of development models is to identify potential
challenges before they emerge or before they get worse. It shows how potential challenges, like
lower tariff revenues, will be overcome within the system over a longer term due to the effects of
other beneficial policies. It is also good, using the model, to test the results of alternative policies
to reverse or mitigate the undesired consequences of the BAU scenario and to mitigate possible
negative effects of the desirable policies. These are essential for long-term decision making of
development policies.
RI was first suggested by CDP experts to meet these challenges. When building and calibrating
the ECOWAS T21 model, it gradually became clear that continued fast population growth due to
the high total fertility rate, would make it difficult to improve the well being of the people in the
region, even with successful implementation of RI. As a result, the FP and RI&FP scenarios
were added.
Are these policy assumptions with numeric values realistic? We are not sure, as no model
projections are by any means certain. But the calibration of the T21 model with the historic
scenario provides a certain confidence that key causal relations are plausible. At present, the
model is being studied and discussed by experts from ECOWAS and its member countries to see
what further refinements and modifications would make the model more representative and
accurate. It is very likely that RI and FP will have effects of varying magnitudes in different
countries.
11
When using the model, the quantities in the above assumptions of the RI scenario, and the future
total fertility rates of the FP scenario, can all be modified by the user. Simulation results of the
different rates will be generated immediately, and they can be easily compared, as was done in
Figure 5.
How should ECOWAS implement these policies and achieve the desired results? This is the
question primarily for humans, including decision makers and all stakeholders. Humans have
the intellect and are creative to discover and invent policies, but when situations get complex,
they are often unable to keep track of all the factors, causal linkages, and feedback loops.
Machines and models are mechanical, but they follow certain niles consistently, even when these
tules become very complex, so they are good for keeping tract of these complex relations to fully
test different policies, see the results, and thus help decision makers reach agreement on more
effectives sets of policies.
Historical data for the ECOWAS 1T21 model was from multiple sources, including the
ECOSTAT database from ECOWAS, WDI from the World Bank, Population data from UN
Population Division, energy data from EIA (US Energy Information Agency), and land, water,
and agriculture data from FAO.
Our next step is to build T21 national models for all the 15 member countries, and in the process,
major challenges for each country will be identified, and alternative policies be tested. Based on
our learning from the country models, as well as feedback from the current review of ECOWAS
T21, we may re-visit the ECOWAS T21 model and update it accordingly.
The ECOWAS T21 model discussed here is an aggregate model, in which all the 15 countries
are combined into a single entity. Another ECOWAS T21 model, called ECOWAS T21
Integrated model, will be developed after the 15 country models are completed. That integrated
model will link all the 15 country models, or the major country models, in a single framework to
simulate and test the consequences of RI for long-term development based on the activities and
progress achieved in each country and their interaction within the region. This will take account
of developments at both the country and regional level, until ECOWAS achieves its goal of full
regional integration.
References
1. ECOWAS Vision Document, March 1, 2011, http://ecowasvision2020.org
2. ECOWAS statistical data office, ECOSTAT, March 1, 2011, http://www.ecostat.org/
3. Weishuang Qu, Hugh Morris and Jed Shilling, “T21 Jamaica: A tool for long term
sustainable Development Planning”, accepted by the Intemational Journal of Sustainable
Development and Planning, to be published in Volume 6, Number 2 in May/June 2011.
4. Weishuang Qu, Ted Chu, Gerald Barney and Jed Shilling, “Challenges Facing China in
the Next 15 Years”, Proceedings of the International Systems Dynamics Conference,
Shanghai, China, 2005.
5. FAO, March 1, 2011, http://www.areppim.com/stats/stats_cerealsxworldxyield.ht
12
