Conference times approaching

Exciting times!

A busy month I have ahead, but all good news. I have got a talk accepted at the Eco**2 conference.

Here is the title and abstract of my talk:

Title: Trophic levels and growth in production networks (in collaboration with J. McNerney and D. Farmer)

We present and analyze a model of long-term price and production flow evolution in an

economy undergoing technological change, based on an input-output

economic approach. This model predicts that prices decrease and

production flows increase non-uniformly across industries due to the structure of the

production network. Industries experience negative price returns

that are proportional to their location along production chains, as quantified by a measure that resembles trophic

levels in ecology. Moreover, the model suggests that mean path lengths

would be an important quantity to consider in characterizing the structure of the

network under a dynamics of widespread efficiency improvements.

I will also give a talk at ECCS'14: Complexity Science and Transport Systems.

Here is the title and abstract:

(in collaboration with F. Medda and A. Wilson)

The trade reforms initiated in Uganda during the 1990s have produced major steps

towards agricultural trade liberalisation that have not only improved competitiveness in

the market but also significant increase in price distortion. One of the main problems is the

quality of logistics chain infrastructures which are still poor and play a discriminative factor

for the success of the growth of trade flows.

Transport costs in Uganda at all levels are very high in relation to its neighbouring African

countries. However, the simple investment of transport, particularly on road transport,

through significant investment policies may not reach the expected results in terms of

increase in trade flows.

Methodology

In order to gauge the importance of transport, in this work we consider an agent based

model (ABM) for the agricultural logistics chain in Uganda. The aim of the model is to

determine the factors which could improve the trade market dynamics of Ugandan farmers.

In this context we design the AB to model the flow of goods from farmers to markets, and

the flow of information from markets to farmers. The AB system is decentralized with no

designated command and control, although a further assumption on the topology of social

network among farmers and traders is made. Each agent follows a decision tree in order to

maximize their utility, and minimize the amount of money paid upfront (risk aversion).

Three main types of agents are introduced: producers, trader and itinerant traders, and

exporters, alongside a simulation of production, transport, and deterioration of products.

We introduce also a banking system, government (taxation) together with an average

weather condition (temperature), which is affecting the product deterioration rate. In

particular, we introduce warehouses, in which consolidation of products take place.

Products are transported mostly using bicycles, motorcycles, pick ups and lorrys. The speed

and capacity of these are modelled according to realistic values and actual costs.

Using the bankruptcy rate as an order parameter, we check the effects of the following

policies:

• Raising/reducing transport costs

• improvement of logistics infrastructures

• coordination policies for logistic consolidation centres

We compare the effect of these policies with the equilibrium values of the bankruptcy rate

in the basic simulation.

Data

We focus on 7 districts located in the Central and Eastern regions of Uganda: Luwero,

Mpigi, Masaka, Iganga, Mitiyana, Kamuli, and Mukono. We use available GIS data to define

the spatial, geographical and demographic environment of the agent based system. The

spatial and geographical environment is captured using a network structure divided into

paved and gravel roads, which affects the speed of transport.