My next post should be called ‘Hand Movements and Tribalism’

Focus on the Systems, Not the Politicians


In that last six months or so, articles about politicians on my Facebook reminded me of an important feature most people seem to have. When it comes to politics, people are interested more in people than in processes. Understandable, as ultimately it is in your best interest to have leaders that have similar values to you. However, at this point in time political processes and desired outcomes should be the majority of political discussion. Individual politicians receiving so much attention from citizens is really a symptom of a poorly designed political system. A political system should be be like a bridge, computer chip, or other engineered system, highly robust to variability in individual components, and while the operation at low levels of abstraction may be interesting, they should not be as interesting as what is happening at the system level.

I remember watching part of the 2016 presidential debate and finding it odd that NATO was a significant topic in debates, yet I haven’t seen any data showing that NATO’s influence correlates strongly with economic outcomes. I have also not seen any significant changes in political policies that seem to impact my life as much as technology has. It is obvious that most improvements to our lives are due to technology improvements, not political ones. I’m not sure what this implies about humans, but my intuition tells me it is not a good thing. Furthermore, the exponential change of some technologies means the influence of technology over economics, social issues, people’s lives, is growing relative to politics, which in a stable political environment is probably some noisy linear function.

As a student, I remember an occasion where I had a casual conversation with a professor I admired during a department barbecue. We were talking about various problems in San Francisco and he went off on a rant highlighting the social and infrastructure problems in San Francisco. If memory serves me correctly he told the story of the last time he rode a bus. A mentally ill person on the bus started banging their head against a window, which was when the professor decided to never ride the bus again and instead to drive a car. He then posited that the reason we have so many social and political problems is because politicians are idiots and all the smart people go into engineering and science…etc. I’m not sure how literal that conclusion was supposed to be, but the conversation has stuck with me. As a university student, I also found it appalling that the field of politics was called ‘political science’, which seems like a rash coupling of words. Now I think it reflects the American societies poor understanding of what science is. ‘Political science’ is actually doublespeak that misleads people‘s understanding of the current state of politics and political progress.

Therefore, in order to change the way people think about politics and take political actions, and for politics to have a more positive influence in society, but most importantly, to improve the quality of my facebook news feed, I propose putting more engineering and science into politics.

Create True Political Science

One of the major issues in politics is that it’s not scientific at all. First, let’s get on the same page about what science is. The first Google search result defines science as:

“the intellectual and practical activity encompassing the systematic study of the structure and behaviour of the physical and natural world through observation and experiment.” has additional definitions that may contain the meaning most people ascribe to science:

“4. systematized knowledge in general.

5. knowledge, as of facts or principles; knowledge gained by systematic study.”

If we agree on the these more general definitions of science, then I’m not aware that we have a lot of systematized knowledge about politics or have extracted principles from the systematic study of politics.

Therefore, to develop political science more thoroughly in application, there should be clear systems to study politics and extract principles. To start, many methodologies can be borrowed from other scientific fields. Writing a new law should be like designing an experiment in natural science. A new law is like modifying an independent variable, and therefore the dependent variable should be selected and the methods for measurement should be decided in advance when possible. For example, if a new law, let’s call it Law A is written to ban Drug 9, then the hypothesis is that Law A will reduce the use of Drug 9. If Drug 9 use can’t be measured then that’s probably not a good law to write, but let’s assume it can be measured. By using this methodology, more laws will become more easily falsifiable by putting a burden on both law makers and voters to generate evidence that the null hypothesis should be rejected. In addition, strong evidence will be generated to support useful laws.

This system can be further improved by more advanced experimental design. For example, one trick molecular biologists use is to have a time limit for measuring the independent variable and also a threshold at which to stop the experiment. For example, a molecular biologist may use Green Flourescent Protein (GFP) as a reporter for the expression of another gene that is the independent variable in an experiment. The biologist may plan to end the experiment in seven days and measure the GFP expression by fluorescent microscopy daily. The biologist may also set a GFP intensity (or GFP positive cell count) threshold, and stop the experiment when the GFP is very bright under the microscope. Connecting this to making laws, a laws can be made with both time limits and thresholds under which it assessed for it’s effectiveness.

Engineer a Political System

Instead of proposing sweeping changes to the current political system, I propose starting by simply improving the tools, components, and design of political systems in general.

Political systems are designed using power structure descriptions in constitutions, this is like the computer hardware. Law is a layer running on top of the power structure, like software, and is written using legal language. It could be argued that the tools for designing these systems have not been significantly updated in thousands of years. On a side note, while I think the US constitution is an impressive and well-thought out document, it mixes law and power structure descriptions, which may not be an optimal design. On the other hand, supreme law may have it’s merits, just as in computer engineering it is sometimes necessary to design Application Specific Integrated Circuits (ASICs) for increasing the speed of specific computations that could run on the software layer.

To improve political system design, a Political System Description Language (PSDL) should be developed, similar to hardware description languages (HDL) commonly used by computer hardware/IC engineers. Version control should also be implemented, like software versioning control using Git. A country could have a master branch for the processes and laws at the country, state and local levels. Modules may exist at the state and more local levels, where branches can easily be created and modified. In this case a local power structure change that goes into effect would be similar to a software commit, and in some cases where modifications are highly successful, changes from one branch can be merged into others. All of the above can be implemented while initially keeping the political system the same as it is now.

For the legal layer, syntax and definitions for legal writing should be more standardized. While there are some startups working on the legal language standardization problem, I worry that getting lawyers and politicians to accept such standards will be more difficult then actually creating the standards because a non-standardized system probably benefits them and they fail to see significant but indirect benefits of such a system.


I tend to avoid political discussions with people as much as possible because I have trouble understanding what the goal of the discussion is. Is it a test to see whether we have shared values? Is it a form of therapy? I can only think of one friend, Dan, who I discuss politics with occasionally, and the discussion tends to be about entertaining aspects of politics or politicians, how we could make a financial gain from a political change, or how the political change may effect us or people close to us and what we could do to potentially mitigate negative effects. However, most people that bring up politics with me seem to do so as an emotional expression of outrage, fear or sadness…etc. These emotions may be warranted, but it would be more constructive to discuss how we can improve political systems, by studying politics in a systematic way, and engineering robust political systems with increased utility. An interesting area of discussion would be in selecting metrics that would be appropriate to measure the quality of a political system, and how they would be measured.