Since it’s Saturday morning, and I’m a “night owl”, I’d like to begin writing an exposition “column” here on my blog where I inform people about subjects that interest me — and how they pertain to the subject matter of this blog (political economy/pragmatic right-wing liberalism). These may get very long and start spiraling into incoherence, but I’ll do my best to keep on topic. The in the first of these, I’d like to explain networks in the context of the economy as a whole, and how government works to foster the functioning of these networks. I hope you enjoy, and please voice your opinion in the comments!
Networks: The Web
Networks are an essential ingredient in any complex adaptive system. Without the interaction of many agents, there can be no complex system. Our economic world depends highly on networks with interact with each other in baffling, and often unpredictable ways. Society is filled with networks within networks, and our modern infrastructure is linked by multiple systems (roads, sewers, water, electricity, radio waves, communications cables, etc.)…and each system plays such a vital role that if just a single of these networks failed to function, the entire system would run into incredible difficulty. Companies spend literally billions of dollars insuring their networks with redundancy — but still the occasional blip gets through, and generally ends up costing quite a bit in terms of economic activity.
Despite the importance of networks from an economic perspective, they have not been a central concern of economists until quite recently. Because of the prevalence of equilibrium systems in economics, the study of how networks arise and function, and how they interact has been glossed over. Indeed, if you ever take a monetary economics course, the professor will likely teach the foundations of the money supply and then gloss over the complex interactions which allow $100 deposited in bank A to become $90 in bank B, and so on. This is typical of analysis of supply and demand, and of jokes economists tell each-other.
While macroeconomic equilibrium can be useful as an approximation of aggregate phenomenon, it oftentimes misses some of the most important factors in where the workings of the real economy and the potential for efficiency-enhancing regulation intersect.
Let us take a step down in scale, and look at the nature of networks within a growing business. Imagine that you are the co-founder of a small startup company with only two departments: development and marketing. As head of development you have an idea for a new product. You have a meeting to discuss your plan with marketing, who agrees to base their decisions around your new product, and you are ready to go. Your product is a success and your company begins to grow. Soon, you find that supporting your new product and keeping your books is harming productivity. You decide to set up a finance and customer service department.
Now at this point, your company is (characteristically) disorganized. Your four departments come to you for decisions. As a founder, with your new product idea, instead of one connection, you now have to have meetings with marketing, finance, and customer service to ensure that they are all on the same page. You now have three meetings instead of one. Since you are becoming tired of being the focal point of information, you promote a head of each department, and instruct them to coordinate with each of the other departments (which they do…e-mail inboxes and conference rooms are full!). Now, you have an idea for your third-generation product…but you find something strange has happened. You have a meeting with marketing, but before you get the department’s “okay”, the marketing managers have to meet with finance about a budget. The finance folks have to estimate the customer service impact…and the customer service department needs to check with marketing to make sure their information is in line with brand imaging…you just went from three meetings to ten.
This exponential increase in the amount of meetings is caused by the increase in the density of your network connections. If you added merely one more department, the number of meetings would jump to 25. You have a bureaucratic quagmire. Small changes in one part of your network (marketing) have cascading effects throughout the network. Imagine if one department got held up during the development cycle…every one of your departments would be in gridlock. This is called complexity catastrophe. Densely connected networks become less adaptable as they grow. The existence of such phenomenon is the reason that the armed forces inflict such disproportionate costs on insubordination.
As a firm grows larger, there is a conflict between two opposing forces: as the degrees of possibility within the firm increase exponentially, the degrees of freedom within the firm collapse exponentially. History is rife with monolithic firms which are subsequently “out-competed” by startups with an idea so simple, and so within the power of the large firm to undertake, that it’s often baffling upon seeing it happen. Think IBM vs Dell. Dell Computer, which was founded by Michael Dell with $1,000 of capital, eclipsed IBM within a decade with a simple idea: direct-mail computers. IBM obviously had the means to buy boxes, and postage stamps…but nonetheless saw their retail PC market crumble, eventually exiting altogether to focus on mainframe computing. IBM fell prey to complexity catastrophe. Contrary to popular belief, firms fail not because they are insensitive to change, but because they are too sensitive.
The Federal government often fails because it is an incredibly densely interconnected network that is incredibly insensitive to change — combining the worst of the two precursors of complexity catastrophe. While the government is a very large bureaucracy, it can move fairly quickly when needed (we went to war, after all [degree of possibility] )…however, the fact is, knock-on effects of legislation often creates perverse outcomes on other parts of the network…which then need to be worked out through further legislation, which then…well, you get the idea. All of this is exacerbated by the fact that government is at the whim of public sentiment. Drastically changing the rules every time a new government is in power is highly impractical (although not disallowed), so we almost exclusively deal with slow-moving feedback mechanisms in designing policy.
However, the inability to drastically change the rules is a very strong impediment to the efficacy and efficiency of governmental systems.
The government is in the business of identifying collective action problems, and remedying them as a (theoretically) neutral outside party — indeed, that is the very basis of (sensible) libertarian property rights. To make modern networks work, this is very important. Use the framework above, and imagine contract law governing the operation of the (complex in-and-of-themselves) networks that facilitate the function of the modern economy. Say you want to make a change to your existing communications network infrastructure. Instead of going to the government (as the departments came to you) as the focal point of information about the functioning of the rest of the network, you would have to rely on a densely interconnected apparatus where the bureaucracy would be endless…you’d never be able to make the change. This is the basis of the dictum “possession is 9/10ths of the law”…property rights have traditionally been extremely hard to manage.
But, the government is in a position that cannot be supplanted except by collapse or overthrow. Thus, if a new innovation (like the internet) makes achieving a particular goal much more efficient, the government is stuck in the same position that IBM was in when faced with competition by Dell. A change that is often stupidly trivial can become an insurmountable obstacle. Often, this is what you see…the DMV, for instance. Government which is working well is very hard to detect.
How do we square this circle? Well, it becomes easier with increasing wealth. Factor mobility is much greater today than it ever has been. What we need is jurisdictional restraint. Modern evidence suggests that the “city-state” is a very efficient level of governance…but the “small democracy” is not far behind. We need to break the United States into the American Union, with each state as a separate country (or big cities as city-states, as Arnold Kling suggests). The Federal government would exist to provide coordination of military activity (not operate a standing army), and regulate the union as a free trade/free migration area. Joining the union would entail agreeing to a contracted “Bill of Rights” which each country could not violate, and international coordination problems would be the jurisdiction of the Federal government. Each country would voluntarily pay to support this level of governance. Otherwise, all issues would be domestic to the
This would allow for the efficiencies of competition, and the efficacy of coordination. It would empower people to decide the rules they live by, and would make for a more responsive feedback loop. Canada and Mexico could join as well, under the conditions that they break into smaller countries and agree to the “Bill of Rights”.
Of course, internationally, many of the collective action problems of the 19th century are reappearing…the subject of another blog, perhaps?
Update: Small, aesthetic title change. H/t Billy.