Computational Strategy Formulation for Public Administration: An Agent Based Modeling Case

Presenter: Bashar Ourabi

Part of the Session on Social Systems


1 Like

Model is nicely elaborated. I would thank you very much if the definition of innovation and its quantitative approach is shared. One of the conclusions did not make me too excited about the model which, is related to the competitiveness of countries. Basically, it mentions that high prices of labor, that is wages, would make countries less competitive. I honestly think that countries do not based their sounded competitiveness on low labor costs. Wages should not be considered just as any other price of factors in the economy. It is an end in itself since, well-being of people are inherently connected to wages, income. To provoke greater social utility, other mechanisms can be suggested. Specially, I dont know if model can change results in environments with high income inequality or permanent poverty conditions.

1 Like

Thanks for the presentation. Looking at the code and the documentation it is difficult to figure out what the agency is of the agents in your model. It seems that the only decision they make is whether to migrate. Is this the case? (Sorry if I misread the code). On what empirical evidence is the migration decision based, and are there competing hypotheses?

1 Like

Hi Bashar, thank you for your detailed work on this model, paper, and video. As a newcomer to
ABM, I really appreciate the thoroughness with which you documented your decision making and the literature behind your programming decisions. Like you, the reason I am interested in ABM is its potential use as a policymaking tool. What has your experience been in getting policymakers’ attention with your results and conclusions? For me, even suggesting it as a new methodology in my field of research (the role of intangible assets in the capitalization of nonprofit organizations) has typically led to the criticism that it is not useful because it is “fake data”. Perhaps you are less susceptible to this criticism since you use empirical data for your inputs? I look forward to hearing about your experience discussing ABM results with policymakers. Best regards, Elizabeth Castillo

1 Like

Dear Cesar,

I appreciate your feedback; along with the the model there is a PDF file. In the PDF file there is a section in the literature review categorizing the constituents of innovation from a capital’s paradigm. In the video presentation I make reference to it as three different types of capitals. I then make a link to the types of capital from economic perspectives such as GDP, Market Flex etc… if you feel that i have missed or perhaps I may improve the research document, I look forward to your constructive criticism.

As to the prices of labor I completely agree with your opinion, yet the appreciation of labor wages levels observed were significantly higher in transient (Developing Nations) than mature and well structured economies. My conclusion does not in any way make suggestions such competitiveness on low labor costs as you have mentioned, but I am sure you agree that perhaps that the Low Skilled Labor Wages should not and usually do not exceed High Skilled Labor levels wages. This emerged behavior in some policy settings was construed in my conclusion as noncompetitive only if the appreciation is aggressive and also not in accordance with other skill levels with in the same country. In the appendix of the PDF document you may see that in all scenarios of different policies setting wages did appreciate and that was not construed as a negative effect.

As to income inequality remark you made, the model is not to be mistaken for a general economic model, as I the trick is in making a distinction between a GEM and work that conforms to my original scope as you may appreciate.

Last but not least, The validation exercise discussed in the lessons learned and future work recommendations, in that the calibration of units was not conducted through this work, which would allow for proper validation. I use this research to demonstrate that ABM is a proper tool that may be utilized for Policy Formulation, and perhaps demonstrated a novel way in modeling Policy literature review translation into ABM modeling in a strategic formulation arena. Which I hope you may deem I was successful in :wink:

Dear Elizabeth,

In my neck of the woods, this type of work is unheard of. I had much difficulties selling the computational dimension for policy formulation. Yet I believe I’ve made many steps in the right direction in introducing this novel ABM (Computational) approach. One major selling point, for my environment, was and remains that data for policy is very difficult to come by initially, and this approach actually provides much data for researchers in my environment, providing a span of area to research, in other words a focus. So my approach was proposing direction for future policy work research and also ratios to perhaps to try and later model using conventional econometrics.

I am not sure if my work was the seed for certain policy executions, but in a very short period after my publication, there were two policies regarding permanent residency provisions for high skilled labor in both Qatar and UAE. The Press called it Genius Visas Initiative, you may search for it on google.

As for the fake data narrative, like you, I also did face this hurdle even with Economic Gurus. But after demonstrating a link between empirical data and ABM, and also making a distinction between Discrete System Simulation, econometrics and ABM in my initial proposal I made the case that the ABM would be a tool similar to a compass and perhaps not autopilot. After all economical modelling approaches usually fall short especially in repeat-ability?

I hope my experience may come of help to you. Your title seems very interesting. I look forward to reading it. Finally, thank you for reviewing my work

1 Like

Hello Marco,

The paradigm I try to institute that the patches are actually competing for agents, as when different policy/ies is/are executed the decision for migration is effected causing an emerged behavior due to wealth previously accumulated. In some cases the migration is possible due to cost (in the model there is a friction, distance cost to simulate this) and in some cases it was not, in such cases it was not unemployment increased and competition in wages was eminent.

The emerged behavior of innovation capital generated and time for this generation, in addition to wages for different skills sets were compared at each country (patches) with different initial condition settings to identify a benefit analysis demonstrated in the conclusion table in the report.

The findings were examined to identify a multifaceted policy and also a trigger for different emerging behaviors to consider (Flags) for new policy phase execution. The ultimate hypothesis was to identify the optimized policy mix and phasing to consider by comparison.

I hope this clarifies the ambiguities I regret :blush: