When we talk about a system of government and its structure, as opposed to bills of rights and other explicitly stated policies, we’re looking to optimize the results, but without actually knowing what those results should be.
To a certain extent, decision/prediction processes and value statements can be used to bound the scope of allowable results, but, as we have not identified and implemented even one optimal strategy for these things, the policy of current governments is not fully defined by the constitutional frameworks in which they are supposedly formed. (If we could define and implement such an optimal strategy, the problem would reduce to a question of integrity and experience, which is far easier to resolve than questions of optimal policy.)
Therefore, within the context of discretionary policy authority, we are looking to put people in place that can still get optimal results, but without knowing the exact policy proposals or decision rules that they will employ. However, we do know several things:
– Optimal decisions are most often made when most or all of the relevant information is at least available.
– We do not wish the form of government to become autocratic.
Therefore, we should structure the form of government to prevent the small minority rule, or other such gerrymandered constructs that would make it easy for autocratic rule to take hold.
– When people feel marginalized and unable to affect the operations of government or their daily lives in any meaningful way, there’s a much higher probability that they will seek out and execute alternative means to get what they want. Oftentimes these involve rebellion or other serious unrest.
The most important of these three considerations is the need to get all of the information on the table – the chance of good decisions declines rapidly as information quality declines. Implementing forms of government that reduces the chances of autocracy will improve the consistency of decisions, but does not determine their goodness (e.g. many democratic governments throughout history supported slavery). Finally, the inclusion of various affinity groups is helpful in reducing the unrest and in providing more information; however, if the right decisions aren’t getting made, the inclusion of the affinity groups beyond what is needed to bring in the necessary different perspectives and information is not going to keep the unrest down for long. Ultimately, government has to execute on economic and security responsibilities for it to be of any use.
Therefore, when we consider the manner of drawing districts for republics, we are trying to set them up to get the necessary facts on the table, to avoid unilateral decision-making, and to have stakeholders across the spectrum, in that order. As such, our criteria have to be, in the following order:
– Try to get members of various professions, civil societies, and ages elected to office. The terrain of various districts tends to be a factor as well, so try to draw the districts in a manner that gets people associated with the different types of geographic regions in the country.
– Evenly apportion the people among the districts, so that no single group or small cabal can dominate the government.
– Try to draw the districts so that less-determinative attributes like sex, skin color, tribal affiliation, cultural inclination, and other such traits are present in the representatives that are elected.
Therefore, the first step in drawing districts has to be a breakdown of occupations, any large civil societies or groups, and the geographic features of the country. Then, these have to be constrained to fit districts with equal population. Finally, the lines have to be tweaked to try and diversify along other axes.
Modern societies have a mix of pure rural country, including pastoral and agricultural types, small towns that support the wide open spaces, company towns and regions based around a specific employer, natural resource, or industry, suburban levels of development (what this means depends a lot on the specific country in question), urban low-density development, and urban high-density development. The key factor here is that the bigger cities and suburbs tend to be much more diversified in terms of industries and professions. They are also geographically much smaller, so they have far less chance of spanning multiple geographic zones. Further, the nature of the city economic and social life makes cities much more similar to each other than to the surrounding rural areas in terms of the way people interact with the land and with each other.
I present one such approach to get a good district layout based upon the above criteria: many others are possible based on tweaking the less important factors below, such as from what area the apportioning begins. Population-density-only can be done, but makes it much more difficult to push the districts back into being compact without skewing the likely voting trends.
Start with the least densely populated (by square mile) point in the country.
Widen a square around that point until it hits a change in land use, at which point it takes the shape of the boundary, or until the required number of people in the district has been added. In some countries, there may be enough open, and different types of, land, so that there is both pastoral and agricultural activity in sufficient amounts to make at least one district of both types, in which case this distinction should be considered a boundary; however, in some countries, there is not a clear distinction between the two, so the extension of the square should continue.
If the area still does not have enough people, admit additional land, preferring agricultural/pastoral, small country town, company town, suburbs, low-density urban areas, and high-density urban areas, in that order.
Draw the next district, starting at the point of highest density in the country. Expand its square according to the same rules, but prefer the land types in the reverse order as with the point of lowest density.
Go back to the lowest density spot remaining, then to the highest density spot remaining, and so forth, until a district is formed that cannot make its required population because of isolated spots formed by the previous choices.
Identify the prevalent type(s) of land-use in the non-complete district. Find the remaining un-districted land that has a similar type of land-use, and which is the closest of all such land. Create a corridor to it by exchanging land-use not of the principal type(s) to another district which has more of that type of land-use. This may require three- or four-way exchange to keep the districts equal in population. If no comparable land-use can be exchanged (e.g. there are two smaller cities, separated by a rural resource exploitation industry belt), then create the district without any corridor, composed of the similar types of land-use. (That is, it will be non-contiguous.)
Repeat this final step until all land has been allotted.
(Particularly important for the last district, which will be far-flung, and therefore difficult to travel across.) Get rid of any far-flung districts with lots of small spots by exchanging similar types of land-use and concentrating the population equivalents together, like moving bubbles, into the spot which moves the small spots the least combined distance to join them together. If the spots can’t be moved this way, get them as close as possible, then regardless of land-use, form a district around the spots as close as can be gotten (could be a ring, a star-shape, or others), with incursions into the contiguous territory by the surrounding districts, so that every district still has equal population.
With all districts now equal and roughly of same-land use, swap same land-use between districts, where there are more-determinative differences that are more concentrated if the land areas are exchanged. This is intended to join up similar professions, civil societies, and age groups, such as an industrial park devoted to chemical engineering.
Likewise, after balancing for more-determinative differences, add the most socially explosive factors into consideration, further adjusting the boundaries to concentrate the traits that divide the society. In the 48 states, this would typically be skin color, and in Afghanistan, this would typically be tribal affiliation.
Done! This way you give yourself a good chance to get at least get the farmers and ranchers in with the company men and tradespeople and the professionals and former civil servants all together. The districts that come out of this will tend to lump in the small towns with other predominant types of land-use, but because they are of middling population relative to other areas, when the bubble converge and ring-making happens, they may wind up being grouped together. From a gerrymandering standpoint, it winds up packing constituents from rural and urban areas. The transition areas’ behavior is highly variable based on the particular character of the region being divided.
Shout out for playing purposes:
Daves Redistricting