Investigations into Schumpeterian Economic Behaviour using Swarm

1
Investigations into Schumpeterian Economic
Behaviour using Swarm

• Craig Lynch
• Macquarie University (Graduate School of
  Management)
• Sydney, Australia

2
Structure of Presentation
Background Theory
Research Objectives
Design Approach
Scenario Outcomes
Random Innovation
Risk / Entrepreneur Competency
Predator Behaviour
RD Investment
Imitation Marketing Defense
Analysis Summary
Future Directions
3
Background Theory
Theory of Economic Development (Schumpeter)
Evolutionary Economics
Creative Destruction
Role of Entrepreneur Innovation
Population Focus
Convergence
Strategic Decision-Making by Firms
Industrial Dynamics
4
Schumpeterian Focus
Entrepreneurs
Random Innovation
Risk Availability of Credit
Competitive Behaviour (and Inter-firm dynamics)
Strategic Choices
Predator Behaviour
RD Investment
Imitation Mktg Defense
5
Research Objectives

• Explore economic and competitive behavioural
  concepts (identifying optimum conditions) for
• Random innovation within firms
• Availability of credit and differentiation of
  supply to entrepreneurial candidates
• Predator behaviour amongst firms
• Innovation investment and firm growth
• Imitation and marketing defence investments
  associated with firm growth and survival
• NOTE Optimum conditions represents maximum GDP
  at economy level, numbers of firms, profit
  levels, and persistence of profit (Roberts 2001)

6
Foundation Bruun and Luna Model

• Each agent a consumer and individual artisan
  (producer), at time 0
• Ongoing supply and demand between producers,
  start-up firms, effective firms with workers, and
  consumers
• Consumers purchase via shopping cart concept
  within 2D toroidal landscape, seeking fulfilment
  within 8 neighbours (Moore neighbourhood)
• Entrepreneur evaluates fulfilment history of
  neighbours, leads to start-up firm by hiring
  producers (creating workers), then to establish
  firm
• Account maintained for all agents, used in
  bankruptcy business rule

Agent as consumer
Agent as potential entrepreneur
7
Foundation Bruun and Luna Model (2000)

• Environmental settings
• Bankruptcy level
• Probability of consumer switching from effective
  firm
• Probability that firm will dissolve for
  non-bankruptcy reasons
• Degree of profit sharing between entrepreneur
  and workers
• Confidence threshold (for entrepreneurs in
  evaluating hiring workers)

Sample graphical display of agent types under
different bankruptcy levels
Independent producer
Worker employed by entrepreneur
Entrepreneur with start-up firm
Entrepreneur with effective firm
8
Extended Agent Attributes

• Artificial internal firm attributes
• Product attractiveness level
• Increments as result of innovation, influences
  consumer purchasing
• Financial stability level
• Increments with sales continuity, decrements with
  multiple investment types (predator behaviour,
  RD investment, imitation/marketing defence
  investments)
• Entrepreneur competency level
• Required for credit availability / risk
  assessment by banks, increments based on firm
  start-up formation success

9
Extended Parameters(Probability/Likelihood
Levers)

• Innovation randomness
• Availability of credit
• Dissolution of effective firms
• Predator behaviour ( industry option)
• Strategic RD investment ( type option)
• RD effectiveness
• Imitation investment ( type option)
• Marketing defence

10
Scenario Modules
ATTRIBUTES
LEVERS
11
Generic Execution Environment

• Swarm ABM engine
• 20 x 20 (400 agent) population
• Standard and consistent Bruun Luna baseline
  settings to achieve healthy GDP profile
• 2 time periods
• 0-2000 (with detailed data capture)
• 0-4000 (macro GDP analysis only to determine
  longer-term trends)

12
Scenario Outcomes Stage 1(Random Innovation
Impact)

• Parameter settings
• High, Med-High, Med, Low, V. Low random
  innovation likelihood

• Outcome
• LOW random innovation results in marginally
  higher GDP than baseline and other innovation
  settings

Baseline
LOW Random Innovation setting
13
Innovation Randomness Seed
100 (High)
Innovation Randomness Seed 1,000 (Medium -
High)
Innovation Randomness Seed 10,000 (Medium)
Innovation Randomness Seed 100,000 (Low)
Innovation Randomness Seed 100,000,000 (Very
Low)
14
Scenario Outcomes Stage 1(Random Innovation
Impact)

• No. effective firms highest at low to medium
  random innovation settings

15
Scenario Outcomes Stage 1(Random Innovation
Impact)

• Persistence of profit calculations show LOW
  random innovation firms with lowest levels

16
Scenario Outcomes Stage 2(Risk / Entrepreneur
Competency)

• Parameter settings
• High, Med, Low combinations of credit
  availability effective firm dissolution

• Outcome
• Optimum GDP values when both credit availability
  and firm dissolution settings are HIGH

Baseline
HIGH credit availability and HIGH firm
dissolution likelihood
17
Scenario Outcomes Stage 2(Risk / Entrepreneur
Competency)

• Effective firms formed only in high credit
  availability circumstances

• Maximum no. firms for HIGH/HIGH combination, but
  lower persistence of profit

18
Scenario Outcomes Stage 3(Predator Behaviour)

• Parameter settings
• V. High, High, Med, Low, V. Low predator
  behaviour likelihood

• Maximum no. firms for LOW predator behaviour,
  with persistence of profit calculations steady
  for all scenarios

• Outcomes
• Similar GDP results for all predator settings,
  for both industry grouping settings

19
Scenario Outcomes Stage 4(RD Investment
Impact)
Worst GDP profiles
VERY HIGH Investment Effectiveness
HIGH Investment Effectiveness
Proportional Investment
Non-Proportional Investment
20
Scenario Outcomes Stage 4(RD Investment
Impact)

• GDP profiles similar with exception of very high
  investment and effectiveness, for both types of
  investment

• Medium investment and effectiveness produces
  higher no. firms earlier, BUT least persistence
  of profit
• Proportional investment more positive for Very
  High Investment / Low effectiveness

Proportional Investment
Non-Proportional Investment
21
Scenario Outcomes Stage 5(Imitation /
Marketing Defence Impact)

• 2 scenarios of LOW defence produce best GDP
  results
• Medium investment defence shows GDP decline
  late in cycle time period

LOW imitation / LOW defence
MED imitation / MED defence
HIGH imitation / LOW defence
22
Scenario Outcomes Stage 5(Imitation /
Marketing Defence Impact)

• Maximum no. firms for LOW imitation / LOW defence
  scenario, AND best persistence of profit outcome
• No significant difference between proportional
  and non-proportional investment
• Firms running in the red

23
Analysis Summary

• Essential to analyse at both macro and micro
  level
• Small amount of random innovation beneficial for
  economy (not excessive, however)
• But not for firm profit
• Mixed results (GDP, no. firms, profit
  persistence) for credit / dissolution scenarios
• Low sensitivity for predator behaviour scenarios,
  but higher profit for LOW setting (marginal)
• Medium RD investment effectiveness optimum
  combination for economy but not for firms
• Low imitation / Low marketing defence optimum for
  all measures

24
Future Directions

• Establish credit creation through entrepreneur
  activity, with imitation swarms
• Establish conditions and logic for business
  cycles (Schumpeter theory)
• Establish production dynamics for firms and
  tiered industry structures (intermediate firms
  and product structures)
• Multiple, concurrent competitive strategy
  execution by firms
• Focus on firm balance sheets and PL