An Operating System for the United States
Decentralization in Real World Government
To move forward with an exploration of the impact and implication of the above it will be of value to take steps backward in our inquiry. It is of importance to understand the apparent progression of communication and cooperation toward the distributed system I outline above, but at this point that it will be of great value to step back and look at systems of government and law that are in place today in lieu of a sprawling brainstorm regarding the possibilities of a distributed government. The previous sections are of greater importance to the understanding of systems of communication, and what seems to be the inherent progression toward diffusion of authority. When we endeavor to discuss the implications of those changes on government it will be a greater value to begin (as I have already begun to some extent above) from exploration of the degree to which we have already achieved a decentralized system of government. This discussion will make tangible the implications of further decentralization, as well as the conversation about the virtues and difficulties of such progression.
The “democracy” of the United States sprung from the revolt of the American English colonies. As colonies of England they were subject to the rule of a monarch (or oligarchy, as it were), and rose in rebellion, discontent with a distant ruler whose decisions were, in their view, increasingly alien to life in the colonies, an ocean away from the seat of power. This alienation is a historical manifestation of one of the core weaknesses of a centralized system. Both in government, and computing, great geographic distances strain a centralized system. At great distances the connection to the main hub becomes increasingly tenuous. Looked at with our previous view to the master imperative of problem solving, this strain on the system is a matter of both infrastructure, and the ability of the central hub to understand the problems to be solved. In computing it becomes more bluntly a problem of infrastructure. Distances cause delays in communication, and those delays hold up problem solving. A dumb terminal, connected to a central hub, begins to fail if the terminal is too far from the seat of computing. In essence the problem is no different from a monarch an ocean away unable to understand the struggles of his subjects outside of his line of sight (or effective reporting.) In both cases there is an inefficiency in communication. That inefficiency results in failures of problem solving, whether to respond to complaints of unfair taxes, or to return the answer to a mathematical function faster than could have been calculated on paper. Those failures in communication, and thus in problem solving, result in discontent in the user or in the governed. In both cases decentralization ca provide the solution. By diffusing a certain amount of authority to the governed, better communication is facilitated, and the system is better equipped to maintain its integrity at greater degrees of complexity and in the case of an empire, distance.
Having cast off the rule of the British Monarchy the Founding Fathers of the United States set out to create what they would call, in the preamble to the Constitution, “a more perfect union.” They were well aware that the form of government that came before them had first driven their forebearers to leave England, and had further driven their generation to revolt. Accordingly, they endeavored to create a form of government that struck them as more fair and more effective than that which came before them. Above all they were determined to avoid the trappings of a hereditary monarchy. Drawing on experience, historical perspective, and the ideas of many a classic philosopher, they built what Lincoln would later call, after what might be deemed the culmination of our own internal struggle with the extents of decentralization, a government “of the people, for the people, and by the people.” I can think of no more succinct words to describe the decentralization of power from a monarchy than Lincoln’s. The effort of the Founding Fathers was to conceive of a form of government in which the power to rule was diffused to the people. In doing so they first debated and agreed upon a written Constitution, a document that to this day makes up the core of our system of government. For all its power and weight the Constitution is nothing more than the inbuilt parameters of a decentralized system, no different from those of a computer networking protocol. It is the skeleton set of rules, rights, and privledges which, provided compliance, allows a more complicated system of problem solving, communication, and co-existence. For, at it’s core that is the function of government, to solve the collective problems of a large scale effort to get along.
The form that this government took, outlined by the Constitution, broadly speaking takes the shape of the decentralized web imagined earlier. While the form has changed over the years from that cleanly delineated in the Constitution, broadly speaking, the decentralized structure matches that web. There is a central hub, the federal government, which controls foreign policy, interstate commerce, war, and other matters that have to do with relations from one state to another. Like the central hub in a computing network the idea is that the federal government is tasked with problems of the greatest importance and difficulty, things that carry enough weight to require a unifying answer, or problems tricky enough that confusion or discord has arisen out of State efforts at solutions. The central hub deals with problems that threaten the integrity of the whole system. Off of that central hub branch the States as nodes of this network to which a certain subset of tasks have been offloaded, ideally those that require an ear closer to the voice of the people. By this diffusion of responsibility and authority the Founding Fathers hoped to avoid the problem of the monarchy. At least as outlined in the Constitution the States are tasked with matters of taxation, policing, and others better handled at the node level where localized problem solving capacity can be better used to address the problems of the people, versus the nation as a whole. The system further diffuses as the States each have their own systems of districts and cities, all then taking on a subset of the duties of communication and problem solving. Since before the Constitution was ratified, and continuing to this very day, the debate has raged as to the balance of power between the federal government and the states. That is, a debate over which subset of tasks, and what degree of control over those subsets, a state should have, and which are to be overseen by the federal government. The trend in recent decades has been toward centralization, and it seems this has bred discontent. Discontent justified perhaps because it does not comport with the progress toward decentralization that natural forces demand. Regardless, it would be difficult to argue that this decentralized, web-like form of government has not outperformed, and outlasted the monarchy that came before it.
For all my mention of discontent it can still be argued that such centralization has been slow, and that this slow pace is a product of deliberate design. Our Founding Fathers, wary of the human impulse to strive for power, did not stop at the mere diffusion of power, but also outlined via the Constitution a system of compartmentalization that would task different branches of government with different responsibilities, set those branches in positions of oversight over one another, and in essence against one another in any effort to wrest absolute control from the others. At this point is it worth taking each branch in turn in the effort to understand their decentralized characteristics, and ultimately how they may they may be impacted by a march toward further decentralization.
The Executive Branch
At the epicenter of the federal government is the executive branch. On the federal level The President, on the state level the governors, with tendrils snaking their way into the lives of the people via federal and local law enforcement. In one sense the executive branch is epicenter of the central hub, tasked with the execution and direction of both governmental internal function, and the function of the system as a whole. Through the President and various diplomats the executive branch is also the interface with other system of government around the world. In a computing sense this it is the channel through which other master hubs may interface with the entirety of the system, and it is the node tasked with maintaining the overall integrity of the system as a whole. The executive provides a check on the potential power of other branches, and handles the active maintenance of the system at the street level through the police force. At its most core function it is the in/out port, and the bug screening system.
The system is already hierarchical in structure and thus to an extent decentralized within various agencies and police forces as a necessity of large scale deployment. In a system possessed of the degree of complexity of a government of people there is much to discuss before we will be able to tackle the diffusion of executive responsibilities further. A police force is already composed at the lowest level of individuals, tasked with certain responsibilities, handled on a very local geographic level, literally the blocks they patrol. A president is an important interface and maintenance element. Could this diffused to the people? Maybe through better citizen reporting technology? Perhaps. Would such a system hold up to human frailty? That is a question for another chapter (perhaps I’ll write it next semester,) but if we continue my exploration the roadmap may reveal itself. I expect we will find the answer in the idea of laws as code, as an operating system. When that OS is sufficiently nuanced and specific, it may be possible to diffuse the authority of the executive branch. This would, as with the computing model above, require the difficult, but not impossible, proliferation and implementation of an exponentially more complex system of local, or individual enforcement.
Instead of pursuing a brainstorm of the possibilities of decentralization in this branch I will simply point out that in a more decentralized system of computing for example, the structure of the internet this executive function has been partially distributed, via communally understood protocols (TCP/IP), to the intermediate hubs (ISPs), eliminating the need for a central executive. In a fully distributed system these protocols have been structured such that even the intermediate hubs are not required. Such a system is possible in a networked system of computers solving problems, and to an extent we have achieved this with local government and policing, but even computers require bug tracking and anti-virus software to mitigate the effects of human error and malicious code. At the police level we are beginning to see cracks in the form of poor policing, and citizen unrest. While there is certainly more discussion to be had in this space, the further structure of the overall system is, in a sense built to mitigate and ultimately solve these problems. But the system is slow, possibly too slow for the modern population. So, as with the analysis of any system looking with an eye for improvement, we need to look to the other structures of the system, for the executive is only the gatekeeper, and bug tracker. So who fixes the bugs?
The Judicial Branch
The Constitution established the Supreme Court, and gave Congress the power to create inferior courts. Courts were tasked with the interpertation, and application of the laws. They created, over time, a structure reminiscent of the same decentralized wheel and spoke web you’ve imagined a few times by now: The Supreme Court at the center, with intermediate federal courts, and state courts radiating outward. Each state possesses their own State court system, from a supreme state court, down to local courthouses. Each progressively further diffused node taking on a smaller, but more specifically local set of problems. By problems, I mean specific sets of facts, and the laws that apply to those facts or, in legal terms, cases. If we think of the laws as code, and the cases as specific scenarios to be run through that code, then the factual scenarios that don’t quite match up with what the law envisioned, are bugs/glitches. Because life is complicated bugs are frequent, and occur constantly, and because the stakes are high we’ve developed a well refined system for assessing every single scenario for any mismatching scenario before deciding what in-built solution to apply, or how to modify our understanding of the code to provide a solution. At the core the Judicial Branch, judges, and lawyers as officers of the court included, are tasked only with processing factual scenarios, and providing fair answers as a result of that process. This means, above all, they are tasked with finding bugs, pointing them out with great specificity, and ultimately interpreting the code to fit the facts.
A case starts at the lower levels and goes through a largely standardized, function-like, system of analysis starting from the particular facts of each unique case. For some cases that analysis starts with a showing of the facts before a jury of your peers, for others that showing is before a judge. The Constitution guarantees you a lawyer, and a good lawyer will do their best to argue the facts of the case in your favor. Whether by judge or jury you receive a decision. After that there are available rounds of appeal; in each you have a chance, not to argue about the facts, but instead about whether or not there has been a fair interpretation and application of the law. The Court of Appeals is required to look at the description and details of your case that your lawyer filed, but they don't have to let you come argue it. On appeal your lawyer is trying to argue that the law shouldn’t apply in the way that the judge or jury decided. Here a group of judges listens, then tells you what they think. Sometimes facts are kicked out as unfair or even illegal, sometimes the case is sent back to a lower court to be argued again, but most of the time they just give you a decision. That decision — insofar as as it is an assessment of the application of the laws to that particular set of facts and the law those facts are purported to have violated— is now part of the law. It is now in essence part of the code, as an interpretive layer. That decision will be used later to decide similar facts. It works this way because the courts are also expected to provide consistency, they are expected to adhere to the precedent of their previous cases. Cases are a matter of public record, anyone can look them up, and in theory use the facts of prior cases to make an argument about the facts of their case. This is what they teach you in law school, that the strongest legal argument anyone can make is an argument from precedent. Lawyers spend most of their time looking up old cases and comparing facts; they spend the rest of the time working on ways to help clients avoid breaking a law in the first place. You learn not only, with great specificity, the rules of the system, but also the important complexities that have, over many years, arisen out of the repeated application. The stretch to bring this system parallel with systems of computation is only a slight one. The laws are literally called codes. Lawyers literally use libraries of cases to make arguments deciphering their value, just as programmers use code libraries to sole particular programming problems. Judges process the lawyers' arguments, and control their motion through the system. The legal system is one really complicated, vagary plagued, processing and debugging system for a massively complex operating system we call law. Specific fact sets come in, answers come out the other side, sometimes quickly, sometimes after years and years of consideration and assessment of the code itself. In this sense, within the constrains of application of the code handed down, it is self healing. The whole system is obsessed with fixing itself.
Quirks aside judicial decision making boils down to a system of functions for problem analysis of greater scrutiny. You can kick your set of facts progressively upward to courts of higher authority, each with narrower rules, and judges with more experience. At all levels of appeal your case is reviewed and accepted based on the complexity of the question of law. If it’s an easy question, you’re done. If the court above believes there is a nuanced enough question, they will then agree the hear an appeal on that case. At the top is the Supreme Court, and their decisions impact the outcome of a large number of cases. This means that the question has come up enough, and been decided by appeals courts of the highest levels in enough varying ways that they feel a federal level decision is important for uniform application of justice. The Constitution guarantees that uniform application, and tasks the Supreme Court with maintenance of the system by which that uniform application is maintained. At the Supreme Court level this often means making decisions about what the constitution itself means. This is why we hear about it in the news. Their decisions are big things, and they should be. By this point the analysis has taken years of processing, has been considered by some of the finest minds in our country, and usually pertains to something weighty, like human rights.
Of all branches the Judicial System is the most cleanly decentralized. The hierarchy is very clear, and the progression through the hierarchy is strict. By necessity the fairness of such a system is important, therefore the operating protocol set out from the beginning rises to a very computer like level of conformity and adherence, at every level failure to adhere may result in no decision, and continued analysis. This makes sense. We literally live and die by this system. The Constitution lays out which crimes, or civil cases will first be tried at what level, and sets forth rules to guarantee uniformity and justice. The system is complicated, and slow, but it is hard to argue for any means of speeding this system up, or further diffusing this system. In a computational sense, this system is the Central Processing Unit for the code of our government, and it is built by the complex networking of minds trained in law schools around the country, and narrowed to their best by judge selection processes. In a sense, just like processors, this system is bound by certain almost physical limitations. Beyond a point it cannot go any faster, or get any smaller without producing an unacceptable number of errors.
Similar to the executive branch, further ways to diffuse this system are possible, but they tend to have problematic consequences. Consequences foreseen by the Founding Fathers, particularly Alexander Hamilton and James Madison as they discuss the "mob" vote of Athens that condemned Socrates to death.1 Certainly mechanisms for distribution of authority could be imagined, but in such a system the error tolerance is extremely low, and the learning curve is high. As Oliver Wendell Holmes, or Jerome Frank might point out, the pedigree of a judge is important, just as important as the pedigree of the law itself.23 This may boil down to a very long way of saying that when it comes to the judicial branch, the Founding Fathers of our country whose this structure with deliberate intent. Further decentralization has been tried, and failed to result in justice and cooperation. The development of the US legal system through the centuries, if not millennia, very much tracks developments in computing. First it was of great importance to refine the processing system, to make it smaller, faster, and to increase the fidelity and quality of problem solving on the lowest level. This progression continues in computing, but we have reached a level where processor speed does not limit our ability to solve problems, instead a lack of tailored software to properly present, and handle those problems is the limitation. In our legal system we have built, and continue to refine, a very advanced processor, now we need better code.
Lawmaking in a Decentralized System
The very first section, of the very first article of the Constitution of the United States establishes a Congress, but before it even establishes that Congress it gives it the power to make laws. "All legislative powers herein granted shall be vested in a Congress of the United States." Before the President, before the Judiciary, the Founding Father's saw it more important to answer the question, 'who will make the laws?' The entire first half of the US Constitution deals entirely with Congress, made up of a House of Representatives and a Senate, their jobs, and the process by which they are selected. State constitutions track similarly in the creation of state legislatures for the establishment of state laws. These bodies are in essence councils, made up of representatives elected directly by the people. Those representatives are trusted with communicating the problems and needs of the states, the districts within those states, and working with other representatives to provide solutions through the creation of laws and regulations. In some form, at every level, proposed solutions are drafted, revised, and eventually presented to the council for a vote. Those that are voted in, and signed off on by the Executive, become law. This means even the other branches have to follow the new rule.
In theory, being the operatives of government with their ear closest to the people, these representatives are best suited to present the problems of the people, and propose solutions. The voting and ratification systems are complex in themselves, and structured to prevent the passage of laws willy nilly, with checks in both the executive and judicial branches in the form of a Presidential Veto, or a judicial finding of unconstitutionality. Still, at the core, these other branches are only in charge of oversight, execution, application and interpretation. The only ones who get to present new solutions to problems are those in Congress.
A law begins as a bill, which is just a proposed solution to a problem, something a law might fix. That bill was thought up by a representative in response to something they view as a problem, or a question. Something that will, if answered, ideally help make the society under our government better. At their core all laws are simply an answer to a question, "If someone does a certain thing that society would rather they do not do, what should happen?" If a certain set of facts exists around the setup of a business, how should they be taxed? How fast is safe to drive? And on and on to varying degrees of complexity. Most criminal laws are fairly straightforward, don't steal, don't kill. Civil codes get tricker, but at their core they are still just the effort to provide an answer to a question that has, or is expected to come up. All answer certain sets of facts, just as I laid out above, the same facts that the Judicial system is looking at as they process these codes. Congress is in charge of making these codes, and whether Judges agree or not, whether police agree or not on a personal level, this operating code rules all. These are the functions by which the behavior of humans living under a government are processed, it is the software that the Judiciary processes, and the Executive applies.
The Executive is holding people to the code, and catching bugs, the Judiciary is processing cases and debugging, but Congress is writing the code, or their staff, at least. Most bills are dreamt up by representatives, the product of a story in a newspaper, or an issue raised by someone they represent. The representative brings the idea or the problem to their staff with direction to write up the specifics of a solution, and that staff gets to work on the actual writing. At this point most of the actual coding is done by lawyers, and others familiar with the rules of a system. Once the bill is ready, the representative takes it to a meeting of congress and presents it to the representatives from the rest of the country. They all get a copy, and it is looked over, or not, by the other representatives and their staff. The other representatives will often have changes they'd like to propose, and have had their staff draft a version that includes those changes. Sometimes those changes are wildly off topic relative to the original problem that bill sought to solve. This back and forth usually takes place for a long time before a vote in fact takes place. This is most of what the representatives actually do, argue among themselves about which solution is the best, each representative carrying the power of the people who elected them to office. When the arguing is done, and the votes are cast, we have a law, we've written code.
The process by which this code is established is, like the other processes described thus far, hierarchical, and to an extent decentralized, but the decentralization is messy in comparison to the Judicial, and has only gotten messier. Yes, there is no king at the center, but is difficult to see the central hub down to intermediate node structure we've been able to imagine for the other branches. The laws established by Congress stand apart from the law established by State Legislatures, although the Constitution establishes that federal law is the supreme law of the land. Originally there was greater separation between the duties of the states over their people, and the duties of the federal government. The system was more cleanly decentralized as a certain subset of responsibilities, of problems to be solved, were left to state law. In this way there was a clearer filtering procedure for laws. States were a proving ground. If there seemed to be many answers, and great confusion, then Congress would take action to develop a federal standard. Originally they were more reluctant to do so, but this has eroded and the mess is tangible. We see it today in the form of drug laws as some states legalize marijuana in the face of continued federal prohibition. Politicians say that they are "waiting to see how legalization plays out on the state level," a vague allusion to the way the system is meant to function. But they say this as the behavior of any state legalizing is still technically federally illegal. That is to say, in a realm that is meant to be the purview of the states, a citizen in full compliance with those laws, could be scooped up by the federal executive and jailed. Sometimes this federal supremacy is warranted, sometimes it's simply a mess. Either way, this is an apt demonstration of a system not cleanly decentralized.
Lawmaking is decentralized in the sense that the representatives are chosen by the people, and those representatives on their own come up with the code that will proposed, but there is no clean system of diffusion like we see in the judiciary. In theory, at the most diffused nodes are meant to be the people, but here we have a break, mostly out a necessity. The people vote for a representative, and after that their connection to the making of the code that will then be used to govern their life is mostly finished. There are city and town councils where we can make a showing and have their opinion hear, even state legislatures, and in extreme cases congress, but none of those are bound to do anything but listen. Similarly, the next level up the chain has no obligation to listen to the decisions or concerns of the council below. The level above takes a look at what they're up to, but generally just does what they want, sometimes with motives toward helping the people, more often in recent years with motives more bent toward the security of their jobs, or a "victory" for the political party they represent regardless of the needs of their constituents. There is a reason recent congressional approval ratings are historically low. The system by which we generate the operating system for our country is a mess.
To relate the structure to the computational mechanisms I have thus far used, our congressional system is closer to the distributed model, and at a very very low level. Rather than a central hub, that is, a single person, like a monarch, tasked with making decisions regarding which laws are to be instituted, the Founding Fathers took that responsibility and distributed it collectively to the representatives of the many states. From there the closest to following a web like structure is then the minor respect given to state legislatures, and the tenuous link to the people by virtue of election. The mess we see today is no surprise, the monarch-like center replaced with a room full of arguing representatives, but this structure is necessary. Much like the complexity required in the Judiciary to account for the complexity of life and human nature, the distributed structure of this central mechanism is of great importance. While the inefficiencies may currently bog down public impression of the efficacy of congress, a distributed power structure at this level is the only way to assure laws are passed which best meet the needs of a large, varied and complex populace, and they pass with at least the opportunity of influence from all other parties.
While it is messy, Congressional structure, or some nearly akin to such structure is ideal at the top of the chain by which the operating code of our society is written. That code is intended to represent the will of the people, and the will of the people is meant to be the sword that the representative shall weird, and with a listen to only the quickest of Congressional debates we can hear them do so. Yet approval ratings continue to plummet. Those numbers mean to be a metric by which to measure how well Congress are doing their job. So it seems, progressively more poorly over recent years, has Congress actually represented the will of the people they are tasked to represent.
Thus we reach the point of failure in the march toward decentralization as applied to our form of "democratic" government: the link to the people. It is from this point that I endeavor to envision the application in government, and specifically lawmaking, of technologies developed in the context of computing. These technologies developed not only for the collaborative maintenance of massive bases of computer code, but the system by which coders are educated and able to participate in the mechanism of problem solving. By the application of these technologies at scale, it may be possible to continue to build upon the idea of our Forefathers in their establishment of a distributed system for the powers of Congress. It may be possible to further decentralize the most important, and most powerful branch of government, pushing further toward the people the power over the problems dealt with, and the solutions eventually turned to code, and processed through the system at large.
1. The Federalist Papers, No. 55 ↩
2. Holmes, O. W. Jr., "The Path of The Law," Harvard Law Review, vol. 10 (1897), pp. 457-68 ↩
3. Frank, Jerome. "Legal Realism," from Law and The Modern Mind (Doubleday an Co, 1963) ↩