The final cut on the coalitions & democracy arc:
There are two ways in which coalitions change things - first of all, they are only effective if they bind the behavior of the parties entering into the coalition, or "new party".
This points towards a key flaw in opposition - instead of acting and moving decisions forward, they peacock and posture. They consider "optics" and bray like jackasses, they do anything but constructive work - after all, anything moved forward would be a possible "point" for their arch nemesis, the sitting government.
This suggests that there is a huge reward for politicians when they grandstand - they find this behavior so rewarding that they must enter a contract in order to prevent this!
While this reflects poorly on the current situation - the system has evolved where the opposition does nothing but whine and bicker - it is a plus on the coalition side in terms of expediency: the general stance of a coalesced party will be similar to the previous parties (or, more accurately, the particular representatives/proxies sitting in government). However, note that the "evolved" norm that says it is okay to be an ass and sacrifice the preferences of ones constituents for ones personal gain and pleasure in being an ass could also be more directly confronted: after all, democracy mandates that representatives represent the interests of ones constituents.
The second, and final point, I have on actual consequences: political decisions are "pruning" decisions, they most often address common actions where we take one action, pruning off all other possibilities. When a new party/coalition is formed they will make a new platform, pruning off many of the particular planks to form a coalesced "average". This is key: once pruned, the particular choice possibility on the tree of choices is pruned off - it is possible to bring them back, but unlikely. So this is the an effective change: by pruning the tree early this limits possibilities. In process management (i.e. process, comp. sci, project management, etc.) there is a key rule of thumb: delay all binding (pruning) decisions until the last possible moment, often stated as "don't optimize too early". This keeps one from mistakenly putting efforts into the wrong direction, and allows choices to be considered with all the data that one uncovers as one works. A particular pruned choice branch may be the sole reason someone voted for a proxy, and keeping options open is inherently good.
So this is the lowdown:
- Coalitions are fraudulent, and thus puts the governments legitimacy (and the particular system as a whole) into question. This is, at best, not good, and at worse - dangerous to a functioning wealthy society.
- People - both voters and politicians - believe coalitions have effective consequences
- One consequence is a reflection of selfish behavior on the party of politicians, which is actually against their entire mandate: the politicians must bind themselves in a contract in order to prevent each other from being chumps
- A second consequence is early pruning. While "expediency" is thus ensured, going in the wrong direction fast is not a good thing: better to be a bit less expedient and actually consider which option is best - once made societies resources as locked-in, resources that may be better used.
Thus ends my initial thought arc on coalitions. I find them fundamentally undemocratic, and it would be healthy to actively prevent them - forcing a general election in order to get a "mandate" from the electorate.