A Very Short Introduction to
Agent-Based Development Modelling
Dr. Brett Parris
Senior Economic Adviser, World Vision Australia
Research Fellow, Dept. of Econometrics & Business Statistics,
Monash University, 8 February 2008
What are Agent-Based Models?
 Dynamic computer simulations involving
interactions between discrete
heterogeneous ‘agents’.
 ABMs are based on object-oriented
computer programming: i.e agents are
‘objects’, encapsulating both attributes
(data) and methods (actions).
 Agents can represent anything: people,
firms, governments, land types, pathogens.
 Agents interact with each other and their
environment according to rules which may
themselves evolve.
 The system evolves dynamically – it need
not converge to an ‘equilibrium’
 ABMs can be non-spatial (a ‘soup’) or
spatial – naturally incorporating real
Geographic Information Systems (GIS)
data or realistic network structures.
 Models run thousands of times to get
probabilistic ‘landscape’ of outcomes.
Handbook published 2006
23 chapters
Where are ABMs being used?
 Ecology
 Geography
 Epidemiology
 Political science
 Anthropology
 Economics
 Finance
 Innovation and organisation theory





Combat simulation
Terrorism research
Peacekeeping
Transport & logistics
Operations research
… and combinations thereof.
 Hundreds of papers now published – many in top journals: Nature,
Science, PNAS, The Lancet, even AER, Economic Journal.
ABMs naturally lend themselves to multidisciplinary
studies since they can seamlessly integrate the social,
political, legal, economic, environmental, geographic and
epidemiological dimensions of development.
Modelling economic development
 I am particularly interested in whole-economy models
& models which are used to guide policy & explain how
economy works: i.e. why some policies should be
adopted in real world & not others, & in what
sequence.
 Two broad approaches to modelling economic
development processes: top-down & bottom up.
 Critical difference is where they start.
Top-down modelling
Bottom-up modelling
Different kinds of models
 Closed-form analytic models are not the only kind
of models scientists use.
 For some systems a deductive axiomatic analytic
approach (like mathematical philosophy) is not
appropriate.
 Eg. Geology – models (maps) of earth composition
& structure based on:
–
–
–
–
Detailed field mapping of real world systems
Informed by extensive taxonomies of rock types
Observations of equivalent processes elsewhere
Lab experiments on effects of temp, pressure etc
What kind of system are we dealing with?
Neat separation of micro and macro causal networks?
 Homogeneous agents
interacting with same
strengths
 Complete causal
network at macro-level
 Neat separation of
micro & macro analysis.
 Analytic or
statistical mechanical
approach OK because
system is ergodic – no
one state has any
higher pr than
another.
Or is it more like this?
Interleaving of micro and macro causal networks
“Well, I’ve got to tell you: I’ve never really understood macro.
What I mean by this is that my idea of understanding is having a
model that captures what is going on. In macro we don’t have
that; instead we have empirical generalizations, and those
generalizations tend to break down quite quickly.”
- Ken Arrow (in Colander et al. 2004).
 Heterogeneous,
agents interacting with
varying strengths &
even asymmetrically.
 Emergent macrovariables influencing
micro-entities
Eg. business confidence,
“irrational exuberance”,
interest rates etc.
 Incomplete causal
network at macro-level
 Analytic or statistical
mechanical approach not
OK – system non-ergodic
- have to integrate too
many equations of
motion.
Why agents?
Many assumptions of analytic models are not
appropriate for development modelling. E.g.
– Perfect information & foresight
– Representative agents (eg. single HH for country), ignoring growth &
development of children
– Infinite computational capacities of all agents
– Complete markets & networks for goods, services, capital, risk
– Costless redeployment of labour
– Zero corruption & costlessly enforced contracts & property rights
– Comparative statics rather than evolutionary dynamics
– Irrelevance of money & finance (eg. barter models of trade)
– Use of optimisation over real number field (R+) rather than integer
optimisation of prices & quantities. Real optimisation is not a good
approximation for Diophantine (integer) optimisation problems. It
cannot be known in advance whether given Diophantine problem has a
solution in integers (Hilbert’s 10th problem, proven in 1970 that
there is no solution.)
Why agents?
 In making a simple analytic model more complex, it is often not
possible to relax enough assumptions simultaneously and still
remain tractable.
 Model has to be simulated – either equation-based or agent-based
simulation. Agent-based models can relax more assumptions.
 The solvability of mathematical models
Source: Keen (2001, Table 12.1, p. 265) adapted from Costanza (1993, p. 33).
Type of
Equations
Linear
Nonlinear
Equations
One
equation
Several
equations
Many
equations
One
equation
Several
equations
Many
equations
Algebraic
Trivial
Easy
Possible
Very
difficult
Very
difficult
Impossible
Ordinary
Differential
Easy
Difficult
Essentially
impossible
Very
difficult
Impossible
Impossible
Partial
Differential
Difficult
Essentially
impossible
Impossible
Impossible
Impossible
Impossible
ABMs & Parameters
 A lot of statistical & econometric work required for
ABMs, in data preparation, parameter specification &
output analysis.
 Verification & validation of ABMs is an active area of
research - eg. best approaches to sample over possible
parameter space – Latin hypercube sampling etc.
 “[N]umerical errors can be reduced through computation
but correcting the specification errors of analytically
tractable models is much more difficult. The issue is not
whether we have errors, but where we put those
errors. The key fact is that economists face a trade-off
between the numerical errors in computational work and
the specification errors of analytically tractable models.”
Ken Judd (2006) Handbook of Computational Economics, Vol. 2,
Agent-Based Computational Economics, p. 887.
Estimating parameters
 The parameter estimation problem still exists for tractable models – but it is
often dealt with by arbitrarily assigning values of 0 (non-existent) or 1
(perfect) with standard deviation always assumed to be zero. Eg:
• Ratio of wealth to wellbeing = 1
• Agent’s rationality = 1
• Cost of evaluating choices = 0
• Agent’s info processing capacity = 1
• Firms’ barriers to entry = 0
• Prevalence of mental illness = 0
• Prevalence of addictive behaviour = 0 • Prop. of capital employed = 1
• Mobility of K between countries = 0
• Spatial heterogeneity = 0
• Accuracy of expectations = 1
• Spatial separation of markets = 0
• Cost of redeploying L = 0
• Cost of travel between markets = 0
• Prop. of agents able to access info = 1 • Rate of skill loss of unemployed L = 0
• Degree of corruption = 0
• Info search costs = 0
• Time required for consumption = 0
• Learning costs = 0
•
•
•
•
Heterogeneity of preferences = 0
Rate of change of preferences = 0
Prop. of contracts enforced = 1
Cost of contract enforcement = 0
The assumptions of tractable models
are assignments of parameter values.
These arbitrary values are no more
scientifically valid than the
estimations required for ABMs.
Often less.
Parsimony & Ockham’s razor
 A tension in model building process: a certain irreducible number of
system components and interactions needed to not misrepresent system.
 Eg. Tractor: how many components needed (and at what level of
aggregation?) before we can say we have a model of a tractor rather than
a car or a lump or iron? Depends on purpose.
 Models need clear purpose.
 Models should be as simple as possible but no simpler. (Einstein)
 It’s Ockham’s razor, not Ockham’s glue. Direction of movement is cutting
away from the more complex to the more simple – while preserving the
results. If we’ve never studied complexity of system, & cut it back,
preserving results, model is more likely to be simplistic, not simple.
“I would not give a fig for the simplicity this side of complexity, but I
would give my life for the simplicity on the other side of complexity.”
- U.S. Supreme Court Justice, Oliver Wendell Holmes Jr.
Conclusion: What can ABMs contribute to
modelling development processes?
 Naturally evolutionary dynamics
 Seamless integration of economic, political, social, environmental,
geographic & epidemiological dimensions.
 Heterogeneous agents (age, gender, rationality, health, education etc)
 Taxonomies (code libraries) of market types, behaviours, institutions,
networks etc., based on existing data, field observations and lab
experiments.
 Spatial dimension – linked to real GIS, land-use data, epidemiology
(malaria zones, etc)
 Localised &/or costly information & genuine uncertainty
 Bargaining power and price jumps in value chains
 Explicit modelling of children’s growth & development.
 Natural modelling of integer problems facing economic agents.
ABMs & Policy Work
 A lot of concern among complex systems researchers with
validation & verification of agent-based models (ABMs) –
quite right.
 As a consequence – a view that ABMs shouldn’t be used to
inform policy – quite wrong IMHO.
 Have to look not just at gap between reality and ABMs, but
at gap between reality and current models already being
used in policy work & also difficulties facing analytic models.
 Q not: Are ABMs ‘accurate’? But: Can ABMs offer an
improvement on current models? Even if more humility?
So …
Yes – models need clear purpose
(in this case a model capable of shedding light on effects of
trade liberalisation and aid programs in certain sectors)
- models should be as simple as possible
But … a tension in model building process: a certain irreducible
number of system components and interactions.
Eg. Tractor: how many components needed (and at what level
of aggregation?) before we can say we have a model of a
tractor rather than a car or a lump or iron?
Depends on purpose … Bottom line:
“Models must be as simple as possible but no simpler.”
Q: If we wanted to use an ABM to avoid some of
these problems, what might it look like?
Emergence vs.
accurate time slice
 Here we are not interested in starting with a very simple
system and seeing if it evolves complex emergent structures
& behaviours.
 We are wanting to take a time slice through an existing
complex system.
 Means the system components, relationships, rules of
interaction & evolution must all be in place so that system
evolves sensibly from t0.
 Lots of statistical/econometric work in the background to
get probabilities as right as possible so that model is set up
properly and evolves sensibly.
Bargaining power is fundamental
to economic systems
 There is no Market – but spatially and temporally
separated markets with varying degrees of interconnection.
 Prices/value not just determined by supply and demand
but through bargaining by buyers & sellers
 Asymmetric bargaining power between transactors and
along value chains permits jumps in value in the system –
there is nothing equivalent to conservation of energy in
economic systems.
 Rural Development: Putting the Last First, Robert
Chambers (1983): highlighted the lack of bargaining power
of the poor, due to their low levels of literacy and frequent
sickness.
 ABMs can implement these insights
Location of buyer, LocB
Cost of searching for
alternative seller, CS
Start with
conceptual map of
bargaining
dynamics
Age, A
Education /
Skill, EdSk
Health, H
Income, Inc
Capacity to wait &
search for alternative,
CapW
State of
finances, Fin
Knowledge
of local price
histories, KP
Effective number
of buyers, EB
Knowledge
of opponent’s
situation, KO
Buyer
Quality of
infrastructure,
InfQ
$
Q
Availability
of substitutes
for product,
ASubs
Location of
seller, LocS
Capacity to delay
sale & wait for
alternative,
CapW
Expectations
of product’s
future value,
FutEx
Knowledge
of true cost &
quality of
product, KP
Perishability
of product,
Perish
Knowledge
of opponent’s
situation, KO
Seller
Absolute number
& location of
other sellers, AS
Effective
number of
sellers, ES
Credit rating, CR
Reputation for
prompt payment,
RepP
Absolute number
& location of
other buyers, AB
Gender, G
Kinship, K
Ethnicity, E
Caste, C
Willingness
to lie, WL
Access to credit,
AC
Net wealth, NW
Expectations of
product’s future
value, FutEx
Experience,
Exp
Knowledge
of local price
histories, KP
Reputation for
quality products
& service, RepQ
State of
finances, Fin
Net
wealth,
NW
Credit rating, CR
Access to
credit, AC
Income, Inc
Willingness
to lie, WL
Health, H
Education /
Skill, EdSk
Age, A
Experience,
Exp
The buyer’s
bargaining
power
The seller’s
bargaining
power
Tanzania
Area:
945,090 km2
Population:
35.5 million (2001)
GDP per capita:
US$266 (2001)
Illiteracy:
29% of population
Population structure
10,000 agents
generated
according to
Tanzanian
population
structure:
•age (years)
•gender
•location
(rural/urban)
http://www.tanzania.go.tz/ppu/pcomposition.html
Six aggregation decisions
1.Agent scale
2.Time scale
3.Spatial scale
4.Commodity scale
5 & 6. Financial asset scales
 Key question: How far can we aggregate upwards before
losing information critical to model’s purpose?
 Have to be guided by expert domain knowledge.
1. Agent scale
Above individual
level lose gender,
literacy, health,
disability,
permanent effects
of childhood
malnutrition etc.
2. Time scale
How much real time should
a ‘tick’ represent?
Above 1 week, lose sense
of weekly provision/
shopping, progression of
illness, deterioration in
health, economy etc.
How to model behaviour in
weekly ‘chunks’? Eg. spatial
movement?
3. Spatial Scale
How much real space should
a grid cell represent?
Above 1 km lose sense of
local village scale, distances
to schools, health clinics
etc.
4. Commodity Scale
5 & 6. Financial Scales
Summary
1.Agent scale: individuals (‘super-consumers’)
2.Time scale: 1 ‘tick’ = 1 week
3.Spatial scale: 1 cell = 1 km2
4.Commodity scale: 8 commodities
5 & 6. Financial scales: cash only at this stage
 Key idea: Combination of ‘representative region’ and
10,000 ‘super consumers’ (each representing ~ 177
people) to make model production & consumption
volumes ~ 1/20th of national economy.
 Hypothesis: Model’s demographic and growth effects
translate to national effects; multiply volume effects by
20 to get national volumes.
UML Class Structure
Raster GIS Regional Map of Tanzania
Raster torus
217 x 217
1 cell = 1 km2
Map = 47,089 km2
= ~ 1/20th area of
Tanzania
& ~ 1/20th
impenetrable
perimeter
(borders, lake
shores & coasts)
Red = HHs
Blue = Firms
The Person
Class
The
Household
Class
The
DevModel
Class
Household bargaining power
We define:
•H = average household health level [1, 10] (10 is perfectly healthy)
•E = average education of adults in the household [0, 17] (years)
•M = household mobility [0, 10];
•I = household access to information [0, 10];
•BPH = component of bargaining power due to health
•BPE = component of bargaining power due to education
•BPM = component of bargaining power due to mobility
•BPI = component of bargaining power due to access to information
• BP = BPH + BPE +BPM + BPI
•BP is scaled so that it has a range between 0 and 100.
Health & bargaining power
BPH 
1
(1  e
( H 5.5 )
Bargaining Power
Sigmoid Curve for Health
)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1
2
3
4
5
6
7
8
9
10
11
Health Score
y=1/(1 + e^-(x-5.5))
Education & bargaining
power
Sigmoid Curve for Education
1
0.9
BPE 
(1  e
1
0.8
( E 8.5 )
0.6
0.7
)
0.5
0.4
0.3
0.2
0.1
0
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17
y=1/(1 + e^-(x-8))
Tracking the evolution of bargaining power
within household poverty bands
Evolution of Tanzanian population
Evolution of population
not based on an
equation but on
thousands of births,
deaths (old age,
sickness, infant
mortality, maternal
mortality), marriages,
sicknesses &
recoveries in both
rural & urban areas.
Tanzanian population (millions)
Actual and Simulated Population Data
39
38
37
36
35
34
33
32
31
30
29
28
1996
1997
1998
1999
Actual Data
2000
2001
2002
2003
ABM Avg of 10 Runs
2004
8 years = 416 weekly
‘ticks’
Another recent ABM example …
Lim, M., Metzler, R. and Bar-Yam, Y., (2007) "Global Pattern Formation and
Ethnic/Cultural Violence", Science, Vol. 317, No. 5844, 14 September, pp. 1540-1544.
Model predicts conflict zones extremely well based on spatial pattern of ethnic mixing.
For more information …
 Leigh Tesfatsion’s website:
http://www.econ.iastate.edu/tesfatsi/ace.htm
 Netlogo (Excellent for beginners)
http://ccl.northwestern.edu/netlogo/
 Repast (Java or C#)
http://repast.sourceforge.net/