Institutions in an age of distributed systems

By Nonso Obikili, Contributor   |   02 May 2017   |   4:21 am  

It is not news that Nigeria has institutions that do not exactly work well. On most standardised measures of institutional quality, Nigeria scores poorly relative to other countries. For instance, Nigeria ranked 169th of 190 countries in the World Bank’s ‘doing business’ rankings, and the results are just as bad on many other measures.

The institutional problem can be thought of in a simple way: there is the need for some coordination among various parties to solve a problem. The institution can be thought of as the means for solving that problem. If the institution works well, then the problem is solved at minimal cost. Sometimes the institution can solve the problem but at a relatively high cost, in which case we say the institution is not working very well. Finally, in some cases, the institution might not exist at all and the problem is not solved.

For example, in a society where people want to run businesses, it is beneficial to all to have a system that guarantees that business names do not clash and that ownership structure is known to authorities. Or in a society where people want to pick their leaders, it is beneficial to have a system that ensures people cannot vote twice and that votes can be counted and aggregated in a simple and straightforward manner.

The classic way of dealing with these kinds of coordination problems is to set up centralised agencies to coordinate things. With respect to the business name and ownership problem, the classic thing to do was to set up a centralised agency where everyone was mandated to register their business. In Nigeria, that is the Corporate Affairs Commission. Or in the example of voting, the classic solution was to have everyone who wanted to vote register with one centralised electoral commission, which would be responsible for ensuring that people do not vote multiple times and that votes are counted and aggregated in a simple and straightforward manner.

Unfortunately, as we have learned the hard way, these institutions can sometimes not work very effectively and can often create more problems than they solve. Improving the competency and capacity of these institutions as they are, is, of course, necessary, as is currently being done. However, a question that should be on every policy maker’s mind is: if there is another way. Can we figure out a different way to solve these coordination problems? The answer nowadays is “yes we can,” with lessons from computer science.

Distributed systems are a very common and important part of computing today. With the proliferation of internet enabled devices, computer scientists quickly figured out that it was more and more difficult and expensive to build computers that could handle all that data and traffic. Imagine the amount of computing required to deal with the millions of people searching for stuff on google every second? Or all the people liking things on Facebook? To deal with this problem, computer scientists figured out that it was easier and cheaper to have many computers talking to each other, as opposed to one giant computer. The many computers are all capable of operating independently but all perform essentially the same function and share the same database, which is sometimes also distributed.

Anytime you search for something on google you are not interacting one mega computer somewhere but one of many smaller computers. Anytime you upload a picture on Facebook you are not interacting with one super computer somewhere but one of many smaller computers.

The benefit of this type of distributed system is that you almost never have failures. One supercomputer can break down and one small computer can break down but it is very difficult for all the computers to break down at the same time. There is also more room for innovation and the introduction of new and faster computers to the network without having to break the bank.

Some networks, like the blockchain, even take the idea of distributed systems a step further by designing networks that allow anyone to add their computer to the network while maintaining the integrity of the system.

What do all these mean for our institutional problems? It means we no longer have to think about centralised agencies as the only way to solve the coordination problem. We no longer need a corporate affairs commission to solve the problem of guaranteeing business name do not clash and that beneficial owners are known to authorities. Every state or local government can have their own registries as long as they all talk to each other. We no longer need an independent electoral commission to guarantee that people do not vote multiple times and that votes are aggregated quickly. We no longer need things like centralised birth, death, or marriage registries, or central agencies for issuing drivers licenses, or even central agencies for identity verification.

The technology and knowledge is there to design systems that will allow us to change the way we think about institutions and build systems that help overcome some of our coordination problems, the very reason we think about institutions in the first place. Systems that, like the computers, will be more reliable, accessible, and open to innovation and improvement. We just have to be bold enough to take a leap into the future.

Nonso Obikili is an economist currently roaming somewhere between Nigeria and South Africa and tweets @nonso2. The opinions expressed in this article are the author’s and do not reflect the views of his employers.

You may also like