Government by data, a different approach to separation of church and state

In my recent posts promoting the concept of government by data and urgency (here, here, and here),  I described government by data as a very different approach to governing.  In place of a large collection of perpetual laws that each required extensive debate and deliberation, government by data will have a short list of rules based on the best available data and analytics where these rules will be small in number and will expire in a short time.    This new form of government periodically runs algorithms against the available data to produce an updated list of rules to replace the rules that expire.  In contrast to existing government where a large number of laws are feasible because they last long enough to allow people to adapt to all of them, the new approach produces and expires new rules too quickly to have a large number of rules.   To work, the government must keep the active set of rules to be small so people can learn how to comply.   I suggested that the population can learn to tolerate lower priority issues to be without governance so that the number of enforceable rules can remain small.

In my last post, I used the US experience with alcohol prohibition as an illustration of how a short-lived rule might work.   In that post, I also complained about the unfairness of imposing long-term prison sentences for violations of a law that was later repealed.   For government by data, we need to approach justice differently.  In particular, we need to recognize that punishments should coincide with the time-frame of the rule’s period of validity.   It does not make sense (to me) to extract a punishment for an offense that has long become obsolete.   Even in cases such as murder or burglary, many of these may have been a consequence of conditions during prohibition and that may not have happened if the law was not enacted.   I object to the notion of extracting a penalty for crimes that are no longer relevant.

In an earlier post, I made a distinction between accountability and justice.  In that post, I spoke more narrowly about the problem of seeking accountability from a responsible decision maker.   The project of accountability for a decision is to obtain the best information about what what the decision maker considered in making a decision: what data did he use, and what doubts and fears did he consider.  The goal of accountability is to obtain information about the decision making process so we can make better decisions in the future, or so we can decide whether we should keep the decision-maker in that role.    I suggested the project of accountability should be separated from justice.  The goal is not to seek punishment of the decision maker.   The threat of punishment (or seeking justice) over the accounting will discourage full cooperation by the decision maker.   I suggested that for seeking accountability from decision makers we should grant automatic immunity from any justice.  What is important is to understand how a decision was made, and to understand whether the decision maker is competent to continue in that role.   We need to dismiss our desire to seek justice so that that we can improve our ability to make better decisions in the future by obtaining a more complete accountability from the decision maker.

In the more recent posts on government by data, I presumed an automated decision maker.   The rules would come from data processes alone.   The population will have access to view the data for themselves, and they will participate in the selection of algorithms.   Machines will automatically generate new rules based on those data and algorithms at prescribed intervals.

In effect, this puts the entire population in the role of becoming decision makers who have an obligation to follow the rules from those machines.   To improve future rule-making, we want accountability from the entire population.   By analogy to the above example of decision makers, that accountability should be free from expectations of justice.

That led me to observe that government by data, in its purest form, has no use for justice.   This form of government can operate without any expectation of lasting punishments.   Government by data offers a more ideal form of separation of government from religion because justice is irrelevant to government by data.

Much of most governments (especially in the West) derive from a divine delegation concept from monotheistic religion.   The major monotheistic religions sharing the books of Abraham have in common a concept of a very powerful God who pay attention to people individually but grants everyone a free will to decide whether or not to follow his will (and be virtuous).  This concept of a God is one that judges not only the individuals but the societies of individuals in terms of their using their free will to good ends, but a God that does not interfere directly in any way.

The connection to government is that it is government’s role to be the active agent to promote God’s will.  Governments provide the necessary actions in daily life to best assure God’s will.   In particular, any justice that occurs must occur from man’s hands, and ideally from governments.   The project of punishing for crimes has the religiously-inspired goals of obtaining justice, providing deterrence for others, and encouraging reform for the individual’s own soul.

My observation is that this justice is not a necessary feature of government.  In my discussion of government by data, I’m describing data technologies that do not yet exist but may come in the near future.   Currently, we don’t have the volume or velocity of data (and analytics) to accomplish the goals of the government by data.   However, it seems likely that we soon will have that capability.

Part of the volume and velocity in data involves detailed data from every individual.   Data can capture aspects of people’s competencies and dispositions.  While people will always surprise us with new information, the government-by-data can quickly use that data to make future rules or plans to best employ those people in the next iteration of rule making.   The access to data about people’s prior actions is another way to look at accountability.  The accounting of their actions becomes data to make future rules to best use them in the future.

The accounting may involve judicial processes such as a careful trial to determine the facts from conflicting accounts, but the only goal is to obtain data.   At the end of a trial, that conclusion of guilt will become part of the data but that guilty party may be free to go on living within society without further penalty.   The government will use the data of that past guilt to make future rules that affect how that person can participate in society’s priorities.   Such a government gains not benefit from extracting punishment for that crime.

Government by data may have some form of punishment or restraint that may be imposed in the short window when a rule is in force.  The punishment may be harsh but quick in order to reintroduce the disobedient back into society with more encouragement to obey the rules.   Alternatively, the punishment may involve incarceration for the remainder of the active period of the rule so that the remaining population of cooperating citizens can provide clean data about the impact of the rules in force.   I allowed for a third option of a death penalty, but I presumed this would be very rare for the same reason they are rare today: death penalties require an even higher standard of elimination of doubt.

In a government by data, any punishment will need to be quick because there is no sense to impose a punishment after the disobeyed rule expires.  Government by data always focuses on the immediate future with its immediate opportunities and hazards, and with the immediately available resources.   This government demands a high level of tolerance by the population in order to keep the number of short-lived rules small and manageable.  That tolerance will extend to tolerating sharing society with convicts.

In other words, society in this kind of government will be a society that readily forgives crimes.  Society accepts this forgiveness with the knowledge that the criminal has been accountable for the crime, and that accounting now is part of the data record.  Future rule-making will take into consideration the data of all individuals in the construction of new rules and in the selection of priorities.   The goal of the future rule-making is to avoid repeating the conditions that allowed the crime to occur earlier.

This form of government does not have a goal of justice for the sake of justice.   The goals are to make the best use of resources to exploit the latest opportunities or to mitigate immediate potential hazards.  In its most ideal form, government by data will make and select rules solely on available data and algorithms.   Inherent in the monotheistic religions is that the God does not interfere with man’s free will.   Consequently, God does not provide data.   Government by data is a very complete separation of church and state.

I was collecting these thoughts during the week of January 5, when news came out of a disturbing events in France involving a massacre in the offices of magazine Charlie Hebdo.   The unfolding of the events occurred during previous attempts at describing a justice-free government.    Current events challenge the above thoughts about a government by data alone.   In particular, actions that are this brazen and objectionable demand justice.  This justice is needed to defend our concept of society that does not accept that expression of speech could justify murder.

As I mentioned above, there is a narrow window where government by data may impose a penalty.  The penalty may be harsh and it may include the death penalty.  The actual events (as of news on January 9) is that the murderers died in a firefight.   I think this outcome is consistent with government by data.   Government by data has as its primary goal the need to respond to immediate hazards.  Engaging in a firefight with armed combatants can happen in my concept of government by data.

The problem would have occurred if the murderers were captured alive.  Had that outcome happened, there would be an investigation and trial followed by some kind of long-lasting penalty (I understand France does not have a death penalty option).   In government-by-data, there would also be an investigation and trial.   The distinction is that there would not be a long-lasting penalty.  I allow for the possibility of a death penalty which may apply in this case.  The only other acceptable option is a brief punishment that may be harsh and cruel but would not result in lasting disablement or disfigurement.   In the scheme outlined above, if the convicts were not executed, they would be quickly reintroduced into society with a short period to recover for whatever injury they received from the brief-but-harsh punishment.

It is hard for me to imagine it possible that a free society can offer forgiveness for this kind of crime.    But, that is what I’m suggesting must happen in the pure form of government by data.   Government by data only acts on opportunities and hazards that are coming out of the future.   The past is only relevant as a source of data.

Part of what is hard to imagine is the volume and velocity of data that will be available to make this government work.   The government will have data on these individuals, their convicted crimes, their competences, and their attitudes.   This data may inform the government of a future hazard that needs to be addressed by new rules.   The future rules may require doing something specifically about these individuals.   I suspect, however, the data will show that these individuals are not the most urgent hazards.   Instead the data will identify other groups who are more likely to cause a more severe tragedy.   The priority of the government by data is to make rules to address the most likely and most damaging hazards in an attempt make those possibilities less likely to occur or less damaging if they were to occur.

The volume and velocity of data that would be required to make this work is huge and probably far more than current technology can support.   However, if sufficient data technology becomes available, then it may be possible to have such a government and such a society.   It will be a very different government than we have now.   In particular, it would be a government without justice and thus without any link to a non-interfering monotheistic God.


7 thoughts on “Government by data, a different approach to separation of church and state

  1. Pingback: Government by data and urgency redefines crime and punishment | kenneumeister

  2. Pingback: The role of charters in government by data and urgency | kenneumeister

  3. This Philosophy Experiments offers many online quizzes to problem concepts of morality. For fun, I answered the questions from a dedomenocracy perspective, in other words how a data algorithm might decide on answers based on data alone without consideration of morality or larger truth.

    At the end of each quiz is an automated logical or rhetorical evaluation of the responses I provided. Not surprisingly, my responses do not score very well. In particular, it disapproves of the inconsistency and arbitrariness of the choices. According to the philosophical point of view of that site, I need to think more carefully about my morality because my decisions are so arbitrary. In each case, I did make the decision I thought would make the most sense give only the data present without considering morality. The site’s disapproval of my sense morality illustrates my point that dedomenocracy is incompatible with morality.

    The question is which is really a better way to run a government. The site insists that strong sense of morality is essential to government (and this is consistent with my earlier defense of accountable decision makers). I’m not convinced that morality is essential for government. I think morality can lead to bad government such as making choices based on justice for past events instead of relevance for future ones (see the quiz for the fat man for a good example.

  4. I started listening to this podcast and didn’t get past the introduction because it argues that morality is in tension with self interest because it takes into account other’s interests. I disagree. A moral decision may be in tension with the totality of everyone’s self interest. For example, it may be argued that in light of the damage humans do to the planet the most moral act may be for human extinction. As I said, I didn’t get past the introduction but I do intend to listen to the whole podcast later. This is as convenient place to bookmark it as any other.

    In my recent posts, I’ve been exploring a thought experiment of a government that makes decisions purely on objective data and algorithms with no human inputs of prior preconceptions including morality. Although I excluded moral truths from decision making, I have not yet explored the moral implications of a dedomenocracy.

    Can decisions without moral inputs be moral? An earlier post I suggested an answer. In that morality (labeling of good or evil) is something observed after the fact, not before. In this formulation, morality comes as a consequence of the action, not the intention. The tie-in here is the possibility that dedomenocracy with its heavy discounting (or rejection) of moral arguments may in fact be our best option for a moral government. The point I’ve been making in recent posts is that the goal of dedomenology is frequent short-lived rule making that responds with agility to current conditions to take the best advantage of opportunities or the best avoidance of hazards. The justification is the realized benefits will outweigh the damages experienced by the community as a whole.

    A dedomenology may be the best approach that takes into account the interests of the entire community and thus may be the most moral form of government now made available by data technologies.

  5. Pingback: Spark data: distracting data deliberately introduced to influence analysis | kenneumeister

  6. Pingback: Dedomenocracy is both libertarian and authoritarian: punctuated libertarianism | kenneumeister

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s