Analysing and Modeling the Postal Systems in Least Developed African Countries - An application of System Dynamics

Diploma Thesis - Summer 2006

Summary

Problem Description

The Universal Postal Union (UPU) is a specialized institution of the United Nations Organisation (UNO). It has the vocation to stimulate the sustainable development of universal postal services of high quality, efficiency and accessibility in order to facilitate communication between inhabitants of our planet.
The postal systems in least developed countries only exist in a rudimentary way. Missing basic infrastructure, low number of post-items and illegal competition are problems among many others, the postal administrations have to face. The UPU, in coordination with the millennium development goals of the United Nations, initialized different projects envisioning an improvement of postal systems in Third World countries. This diploma thesis is done in this context. We consider the special case of Least Developed African Countries (LDAC).

The main objective of this diploma thesis is to provide new insights into the problem through methods of Operations Research. Hence, a comprehensive analysis of the postal system in one least developed African country, a set of innovative recommendations having in view an improvement of the system and the realization of a model to evaluate these actions, should result from this work.

Situation Analysis

In this part, the main kinds of networks have been analyzed, in order to get a better knowledge of network economies and to identify which strategies have been implemented and if they led to success or failure. The postal systems in a high-income (Switzerland), a middle-income (Brazil) and a low-income country as well as the telecommunication systems in developing countries have been studied. For analysing the postal system in least developed African countries, we decided to focus on Niger.

Postal systems in Least Developed African Countries

The postal delivery system in LDACs is very often based on a post office box system. One has to rent a letter box at a post office to be able to receive mail. Their price is very high and the way to the post office may be quite far. So both sender and addressee have to pay and walk the way to the post office.
The number of potential postal users (persons above 15 able to read and write) is low and sparsely located, because the population density and urbanization is low. The post-item volumes are usually low and falling, so no economies of scale are possible, like they are in industrialized countries. In Niger, the total mail volume is an average of 0.13 letter per inhabitant and year, and 1 letter per potential postal user and year. For comparison, in Switzerland these averages are 760 and 1048 letters per year and inhabitant, respectively potential user. Postal operators are usually public administrations and get important subsidies from the government.

Network externalities

Network economies are based on network externalities, also called club effect. These state that the utility a new client can gain in joining a network is function of the number of persons already connected to the network.
The demand curve usually follows an s-pattern as depicted below:

The growth is flat in the beginning, when only few people are connected. At a certain level of members the club or network size grows very fast. It flattens again when near to the maximal demand.
Postal systems in least developed countries are at the very beginning of the s-curve, whereas in industrialised countries they are almost at the top. There, almost everybody is connected to the postal network. At that point, the curve is very flat, that is the reason why postal operators in industrialised countries usually neglect the club effect.
But in least developed countries it is very different. The growing potential is immense, but for growing the network needs to reach the take off point. Presently, the number of persons connected to the postal network is low, so the incentive for new clients to join the system is low as well. The membership price, the rent of the p.o. box, does not seem to match to the utility of joining the network.

Two-sided market

Postal operations set up a two-sided market. The utility of postal services are different on the reception and the expedition side. The prices have to be adapted. Econometric analyses show a higher price sensibility for reception than for expedition. But the current price structure (p.o. box rent - stamp price) does not seem to reflect the actual market price elasticity.

Modelling

The main objective was to develop a model allowing the analysis of the main variables and their interactions, in order to improve our knowledge about postal systems. We focused on a global model, representing the general dynamic of the system as a function of different politics and strategies. We used the system dynamics method to model the postal system. System dynamics is a tool to design and manage complex systems, which may be containing multiple feedback loops, long delays and non-linear responses to the implemented decisions. Considering the low state of development of the postal networks in LDAC and the lack of specific data, a global model is an essential first step to further analyses. This model will provide a basis for discussing possible strategies and testing different scenarios and their influence on the system over a few years.

The description here is purely qualitative. The main influencing variables and the interactions are depicted in figure 2.
The offer represents the service quality and the prices provided by the postal operator. The needs are the wishes the population has for postal services; they are exogenous to the system. Since we consider a short horizon of time, we assume these to be constant over time. The demand is the expressed needs, the sales. The demand and the prices include the two-sided market problematic. The club effect is modelled through the variable network attractiveness. It is influenced by the number of existing members, the needs and the service quality. The number of new members is then influenced by the network attractiveness and the p.o. box price.

Positive loops and limiting factors

Two positive loops are depicted in this figure:

The club effect: attractiveness --> members --> attractiveness
A large positive loop following the cycle: profit --> offer --> demand --> profit

These loops are limited by the needs and the financial means which have an upper bound. Moreover, we assume that the benefits will be reinvested for improving the offer. Then, at short term, the profit will decrease, due to investment and maintenance costs and prevents the positive loop to take off.
Precise equations have been defined to replicate the relations between the variables. They are exposed in the enclosed report.

Local specificities

The population density and the economic development may differ locally within a country. So the needs the population has for postal communication are different depending on the region. Furthermore, in least developed countries, interregional communications are more important than local ones. Because of the distance to the post office and the high p.o. box price, substitution possibilities are used more frequently for local mail. To integrate these local specificities, the country was divided into three zones (dense, medium and sparse) according to the needs for postal services and some of the variables (resources, demand, needs, ...) were defined locally.

Model validation and limitation

The model is validated by the tests on past data. The model prognostics and the reality are highly similar. Moreover, the different tests we have processed show realistic prognostics which have been validated by the development economist of the Universal Postal Union.
The model is restricted to least developed countries with a p.o. box delivery system and a short horizon of time. We don't consider exogenous variables to change over time. The complexity of the model is also limited, due to the time constraint under which the project had to be carried out.

Application and Results

The model has been implemented using a system dynamics specific software, which is able to run simulation through solving the underlying difference equations. We could then test different scenarios. The price structure, the type and the location of investments in resources and the impact of training have been studied.
The key findings resulting from these tests are the following:

• Universal Service Obligation should not be considered as a constraint, but as an opportunity for development. Increasing the network size using alternative means of distribution may lead to considerable improvements. Closing post-offices leads to a short-term profit, but sets the network in an even more negative trend.
• Changing price structure, that is providing free delivery while increasing the stamp price to maintain the current income, shows extremely positive effects without need for investments.
• The training of postal employees can improve the service quality, but this action alone has only a limited impact. It should be combined with others.
• Financing the necessary investments is difficult since assets are low. A reasonable increase in the stamp price to provide better service quality makes sense.

Outlook

This work is a first step to a new modelling approach that focuses on specificities of least developed countries. In a next step more influences should be integrated to the model. The distinction between commercial and private clients, financial services or telecommunications could be added to the model. A more precise division of the country into zones would also be constructive. Furthermore sorting processes, as they have been analysed in the situation analysis, seem far from optimal. Network topology should be studied more deeply.

In collaboration with:

• Universal Postal Union