A Democratic system should work based on continuous feedback on its performance. Without that it works like a uncontrolled and unchecked system which eventually drifts from its original goals set forth by its founding fathers. Right now in India the feedback system is largely based on input from the media. But, what media provides is news. That has the following problems:
In modern India, we have got other ways of collecting information about government matters, RTI. But while RTI provides a mechanism to access important government related information, it is not really easy enough for general citizen to file and understand these on regular basis.
So what we observe is that we already have several sources of information and mechanism of collecting them in place. Now what we need is organizing those information and performing analytics in a way in which it serves the goal of becoming a perfect feedback path for the working Democratic system.
In order to do that we need to define first what we mean by performance of a Democratic system. When do we say we are performing good as a Democratic system and when do we say we are doing bad. There have to be precise measures using which one can say it is doing good or bad. If bad, how bad. If good, how good. That is a complex question and in subsequent writings I will get in details of my thoughts on this. But at this point what we need to keep in mind is the definition of performance may very among different analysts. The goal of this article is to show that it is possible to create a system where different analysts, who can be anyone from a minister to a school student, can create their own definitions of democratic performance. This has to be done with the following conditions in mind:
We can collect lot of relevant information from various sources. It is very important that we categorize these information into different hierarchical buckets so that people understand them easily. For example, at broad level they can be categorized into buckets of "Economic", "Social", "Religious" etc. Then we can further categorize "Social" into "Literacy level", "Hygiene" etc. And so on. For this article, let me assume a flat structure.
Let us assume there are total Cn different categories of information : C[1], C[2],.. C[N].
Now let us define a general performance function F which works on these Cn categories of information and yields a result. Then we can say, performance P = F(C). This is true for a single regional unit where P can be assumed to be same. But in general we will find P needs to be different from one regional unit to other. For example the performance criteria of a village will be different from the performance criteria of a town. So if there are Rn different types of regions(R), then performance becomes P = F(R)(C).
In a simplistic case let us assume R is not varying. So we can have P = F(C). Now let us assume a lenear formula for F where F = x[1]*c[1] + x[2]*c[2] + ... + x[Cn]*c[Cn]. Where x[i] is a real number (i from 1 to Cn). Then we can write P = X * C.
Analysis
In this model we notice that if:
One can easily see that X itself is not a constant measure. In general each analyst will have their own defintion of X which will be visible and ready for comment and feedback from people. So each P can then be subjected to further analytics to obtain a more reliable nationwide performance index.
So we can see that it is possible to build a customizable system which can constantly provide feedback about the performance of a democratic system. I think such a system can play many more important roles, I will discuss them in further writings. The construction of such a system, specially to create the initial structure, is an enormous job in an already established system. I will try to address those challenges one by one in future notes.
- It may, and largely is, biased.
- There are lot of information, but often the source of those information are not presented in structured manner and hence becomes difficult to validate the information.
- Many times the sources are just not trust worthy.
- Generally media reports news which they consider news worthy. There is a wide area of information which is very valuable for evaluating performance of a Democracy but not news worthy for media.
- The output of media news is optimized for public reading/hearing but not optimized for doing further analysis by computer algorithms.
In modern India, we have got other ways of collecting information about government matters, RTI. But while RTI provides a mechanism to access important government related information, it is not really easy enough for general citizen to file and understand these on regular basis.
So what we observe is that we already have several sources of information and mechanism of collecting them in place. Now what we need is organizing those information and performing analytics in a way in which it serves the goal of becoming a perfect feedback path for the working Democratic system.
In order to do that we need to define first what we mean by performance of a Democratic system. When do we say we are performing good as a Democratic system and when do we say we are doing bad. There have to be precise measures using which one can say it is doing good or bad. If bad, how bad. If good, how good. That is a complex question and in subsequent writings I will get in details of my thoughts on this. But at this point what we need to keep in mind is the definition of performance may very among different analysts. The goal of this article is to show that it is possible to create a system where different analysts, who can be anyone from a minister to a school student, can create their own definitions of democratic performance. This has to be done with the following conditions in mind:
- Creating the definition will be easy enough so that one does not need to be highly computer literate to create that.
- The definition itself, including the data it is analyzing and the results of the analysis, all needs to be in public space which any one can.
- There has to be a public comment and voting system in those definitions so that some of those can naturally bubble up as the most famous ones and eventually will reflect what people really want.
We can collect lot of relevant information from various sources. It is very important that we categorize these information into different hierarchical buckets so that people understand them easily. For example, at broad level they can be categorized into buckets of "Economic", "Social", "Religious" etc. Then we can further categorize "Social" into "Literacy level", "Hygiene" etc. And so on. For this article, let me assume a flat structure.
Let us assume there are total Cn different categories of information : C[1], C[2],.. C[N].
Now let us define a general performance function F which works on these Cn categories of information and yields a result. Then we can say, performance P = F(C). This is true for a single regional unit where P can be assumed to be same. But in general we will find P needs to be different from one regional unit to other. For example the performance criteria of a village will be different from the performance criteria of a town. So if there are Rn different types of regions(R), then performance becomes P = F(R)(C).
In a simplistic case let us assume R is not varying. So we can have P = F(C). Now let us assume a lenear formula for F where F = x[1]*c[1] + x[2]*c[2] + ... + x[Cn]*c[Cn]. Where x[i] is a real number (i from 1 to Cn). Then we can write P = X * C.
Analysis
In this model we notice that if:
- We can create a system where C is available to people via Internet APIs and other medium
- C is updated sufficiently frequently (Because actual real time is impossible to achieve)
- We can create a system where people can easily create X
One can easily see that X itself is not a constant measure. In general each analyst will have their own defintion of X which will be visible and ready for comment and feedback from people. So each P can then be subjected to further analytics to obtain a more reliable nationwide performance index.
So we can see that it is possible to build a customizable system which can constantly provide feedback about the performance of a democratic system. I think such a system can play many more important roles, I will discuss them in further writings. The construction of such a system, specially to create the initial structure, is an enormous job in an already established system. I will try to address those challenges one by one in future notes.
No comments:
Post a Comment