#5PAPERS – Week 1 Day 4

by Nivek K Thompson

1.Reference to the Article 

Rios, Vicente & Garcia, Victor 2015, Can Democratic Innovations Improve the Performance of Representative Governments? An Agent Based Model Approach ResearchGate.

  1. What attracted me to this Article?

This paper looks at the macro impact of democratic innovations – which is what my research intends to cover – using agent based modelling – which I haven’t come across before and don’t know much about.

  1. What is it about? (Problem / Purpose / Research Questions)

Responding to study by Ortiz et al (2013) which found that between 2006 – 2013 social protests around the world focused on ‘failure of political representation of current political systems.’ p1

  1. Where does this come from? (Literature / Theoretical Framework)

They look at historical literature about different political systems that say that selection of leaders based on blood heritage is associated with monarchies, elections relate to oligarchies (which they equate with representative govt) and lotteries with democracy.

  1. What did they do? (Methodology & Method)

Analyse different political systems involving citizens in decision-making using agent-based modelling i.e. theoretical approach rather than empirical. Use ‘Monte Carlo simulations of agent based models that emulate the functioning and structure of [three] different political systems.’ p3 The 3 systems are: traditional representative democracy with 3 political parties, participatory democracy where legislation can come from elected politicians and citizens who are members of the Parliament, and participatory democracy where citizens form a second house i.e. Citizens’ Senate. They also look at the risk associated with different models.

  1. What did they learn? (Results / Discussion)

For the Representative Govt model

  1. multi-party systems out perform two-party ones in terms of legislative quality
  2. egalitarian distributions of power increase legislative quality and lower volatility
  3. party discipline decreases legislative quality and has no clear effect on political stability (p.15)

For the Participatory Democracy model (with citizen legislators in a unicameral system) they found increasing citizen participation beyond a threshold of around 15% leads to a continuing increase in legislative quality up to around 70% after which legislative quality drops. The authors suggest 45 – 50% as the optimal level of citizen participation.

For their final model, a bicameral system with a Citizens’ Senate that does not initiate legalisation, only reviews legislation for the first chamber that is based on a Representative Govt model. They specifically set out to test the consensus threshold required to achieve optimum outcomes, which they find to be 75%.

  1. What did I learn?

Agent based modeling is hard to understand – lots of maths!

This paper was interesting but not particularly relevant to my research. Oh well.