Government Model v.2

[Mark_Everson]

Rodrigo:

Support your take on centralization. I think we can maybe layer a few simple things on it to get the right flavor for feudalism etc if necc.

F. Smith may be right in that we might need a good place to 'plug' local govt characters in, even if things at the local level are fairly simple.

I will be on vacation (and out of contact) soon, so I will just throw out one more point....

Have you thought about special war powers? FE in a democracy during the time of serious threat to the nation usually freedoms are suspended, taxes are raised, social spending is reduced etc. If the people support the war they will generally put up with all this for the good of the 'war effort'. Do you have a way to handle this?

[roquijad]

F_Smith:
I can't give you a good answer on province level govt variables, since as you can see, it's still under discussion. Given Mark support on my crude vision regarding local govt powers, I envision at least a province level tax rate and a couple of variables for military recruitment. It shouldn't be more than that.

Mark:
War effort... I'll see how it can be handled...

[Mark_Everson]

I have been thinking in my idle moments about how to handle the interconnections between the government and economic models. I'm going to do a lot more thinking about this while I'm on vacation. Axi, if you can write up those thoughts that you have been threatening us with on the topic and post them, I will hopefully be able to read them at some point while on vacation.

Anyway, at least in my initial thoughts I have some real problems with the extreme polarization between the labor class and the upper class that was put in to drive conflict in the model. It is the assumption that the labor class controls absolutely none of the capital resources of the society that really bothers me. I think this will put too much conflict in the government model, and also do some bizarre things to the economy... I haven't thought this all the way through, but this may be my last chance to post for a week or so, so I just wanted to say a little bit about where I was headed in my thought process.

Essentially, giving the government and upper class a complete control of the capital in the economic model, and so all the benefits derived from the capital, seems to me for much of the early game to guarantee that the labor class are essentially serfs. Although this was true in a lot of the ancient world, there is also much of the ancient world where it wasn't true. In effect, you are making something Always happen that I thought should be a sign that the player had screwed up, or didn't care about the economy!

So at the point where I am in trying to unify the government and economic models, it appears to me we need to do one of two things. Either let the labor class control some of the capital in the society, or put in a middle class that can do the same. (But even when a middle class exists, the labor class should be able to control some of the capital resources of the society IMO). I am kind of leaning toward us putting the middle class in, even though it complicates the model a bit. One reason, is that everyone already knows the term. And most people know that the economy is generally more advanced as a larger percentage of the population is in the middle class... So IMO we can use it as a gauge of how advanced an economy is (Axi, I know you're not gonna like this )

In a related note, I think the different kinds of taxes that Axi proposed is a good idea, and I had in mind to do something similar. However, we should probably not worry about it just at this point. When things get further along we can start a thread just on taxes, and narrow the list of probably 20 or more types of taxes that will get proposed down to something of order five or so.

Rodrigo, Axi, and others... What do you think about these general points?


[This message has been edited by Mark_Everson (edited June 22, 2000).]

[roquijad]

On Mark's thoughts:

Well, I know I once said something like "adding the middle class would be a pain in the ass" when discussing this model via e-mail with Axi before being posted. I see it now as a easier task if we really want to go that way. Anyway, before taking this road, I'd love to know what parts exactly in the econ model become complex/bizarre with the absence of the middle class. If we can find a way to solve those problems, I think that'd be much more convienient than adding a middle class, because, as you know, when adding the middle class we must add a behavior for it in the govt and riots models, which isn't cheap.

That's the most important difficulty when adding the middle class. What about the pro's? 1)More true-to-life; 2)It can help modeling ending of feudalistic regimes and risings of representative regimes (modern and ancient), because the middle class is by far the most important factor in those processes. 3)It helps modeling migrations from country-side to cities and the modern role of cities as THE major productive units in these times.

Measuring the civ's progress by the size of the middle class... you bet Axi won't like it!! (me neither)
One thing is very very important when discussing about the Upper Class, Lower Class and a possible Middle Class: They must be defined according to their role in society and not according to their (relative) PCI. The latter is the intuitive first choice (it was the one I chose initially), but Axi was very clever realizing the fact that class behavior can be modeled much better and coherently when you instead look at their economic role. I'm saying this only to avoid this discussion going in the wrong way.


As for taxes, I, as a player, wouldn't like to handle more than 2 rates...

[Richard Bruns]

I would like to say that I think modeling the middle class is very important. For almost all of Western history since around 1600, the middle class was an important influence on politics and economics. In the modern era, the majority of the population has been considered middle class.

I don't think it is fair or accurate to put all of these people in the lower class. By the current definition, a neurosurgeon making $300,000 a year and a migrant farm worker making less than minimum wage would both be considered lower class. These people will have very different mentalities, and it would hurt the game to lump them together in a single class.

A good working definition of the middle class is "educated, skilled labor." In the economic model, they would usually provide services. The size of the class would increase as the civ creates more educational infrastructure. They would affect the technology model, creating more RP's per capita than any other class. And in the social model, their mentality and values would be different than the upper and lower classes.
[This message has been edited by Richard Bruns (edited June 25, 2000).]

[F_Smith]

Pardon my nose, but I'm trying to get a handle on this model to prototype it, so ya'll can play with the numbers.

Now ,If I'm off base, please tell me, but the various 'social class' choices are already going to be more than just 'upper class' and 'lower class', won't they? I'm thinking of 'noble class', 'religous class', 'warrior class', (no class -- for people like me), etc.

In other words, 'social class' would be similar to 'culture' and 'religion' -- people objects will have different societal behaviors, each person will be commanded by their own 'Social Class' behavior.

Every turn, the code will run a 'turn()' method in each person's 'Social Class'.

So you can have as many different social classes as you want . . . Does this fit in with ya'lls plans?

[roquijad]

F_Smith:
No! The number of classes is no minor thing. We'll have the Upper Class, Lower Class, Religious Class, Military Class and Bureaucratic Elite Class and maybe, depending on where this thread lead us, a Middle Class. No more than that, no less. We can't afford having just ANY number of classes because modeling their behavior is tough. It will also have impact on the riots model, so adding classes increases the game's complexity several times.

Richard and Axi:
As time passes, I'm more and more inclined to the idea of having a middle class. So, let's talk about how it can be added. I see Axi's approach as the more general way to handle an "economic-political society", where each person earns money through providing some level of labor and through providing some level of capital and having returns for it. But, doing it as general as that gives us as many classes we'd like to have, because it's a continous space of possibilities. Therefore, I prefer to define classes like this:
Upper Class: They provide only capital (land, factories, etc).
Lower Class: They provide only labor.
Middle Class: They provide labor and capital.

When adding the Middle Class in the govt model, only the calculations for Ideologies Attractiveness needs a closer look, so this class would evaluate each economic system as a merged view between LC and UC. No major change is needed in the model. Probably the Income Distribution would have to change and become really two variables (two rates).
The problems I see are two. First, demography. This is an issue still not solved even for the current model, because I wanted this to be an outcome of the econ and govt models interactions. However, adding the middle class makes it a bit more complex, because the demographic share of the middle class shouldn't depend only on the current economic system, but also on the size and tech level of cities, where the middle class really exist. And second, computing correctly a PCI for each of these three classes. I imagine it isn't too hard to do, but I'm not sure...
[This message has been edited by roquijad (edited June 27, 2000).]

[axi]

I don't understand why Mark and Rodrigo are so afraid of my reactions to their thoughts about the Middle Class. Do I look like a fanatic?

Anyway, I have to say that it has so happened that I was quite busy last week and I will also be this week too, since it is the examinations period here and one has to study in the midst of a heat wave (my room has a constant temperature of 31 Celsius - outside it is 24 - 39 ). So it is quite reasonable that I have neither the time nor the energy to think about Clash govt or econ.

I will provide some clarifications anyway, before this issue goes astray, in an effort to simulate the external appearances of the Middle class instead of it's internal structure.

Is anybody familiar with 2-phase systems, such as chemical mixtures or alloys? This is the only analogy I can find for my concept of the clash-class-system , which is more or less presented in my related quotation in the riots model (if you haven't read this, please do - allright Mr Bruns?). Every person's income is a mixture of wages and profits (if wages=0% pure UC, if profits=0% pure LC), and through their analogy he can be placed in any social layer. But in reality, no other class than the LC and the UC is a real class, it is just an alloy. Everyone's political profile can be derived in the same way as his income, since the double source of his income causes every "middle class" member to have conflicting interests. We would only have to use the simple cantilever rule to refer any middle class populationXincome group to a couple of respective UC and LC groups. I think this approach is more programmer friendly, but I am ignorant of such things.

The only problem we would have to deal with is the reverse process, or how to simulate the existance and characteristics of a middle class through LC - UC interpolation and I feel that this is what Mark brings forth. We only have:

1)The mean PCI of the UC and the mean PCI of the LC and also of course their ratio = ID (income distribution) variable, all of them determined from the internal provincial market system (modified by state intervention, either fixed (use of SP) or determined by the player (use of EP)), which will split all surplus value into wages and profits.

2) The historical (given from us, the game developpers) function between consumption vs. investment and income.

3) (possibly, if Mark's econ model isn't incompatible with this) The LCR (Labor to Capital Ratio) for each economic sector, which will tell us how many capital units must be used for every labor unit into production, giving us both a rough estimate of which is the income needed for a producer to be independent (his capital employing his own labor, f.e. a farmer) and also of the advancement of the economy (don't you think so Mark?). It is widely accepted that the capital intensive economies are more advanced than the labor intensive ones. This of course has nothing to do with the size of the middle class but with it's independance; if overall LCR is low as in the advanced economies, it is more difficult to be independent.

What I am not exactly sure of is how are we going to combine all this data to give us a representation of the economic importance of the middle class. I feel that as ID increases and LCR decreases, the size and influence of the "Middle class" would decrease. I also think that a high percentage of private investment over consumption would mean an increase in the demographic shares of people in the upper classes, due to larger profit shares.

I think this is an issue well studied in macroeconomics (I think it is the Keynesian theory), but these are deep waters for me, so it would be helpful if someone more proficient in such matters would explain things to us.

------------------
"In a time of universal deceit, telling the truth is a revolutionary act."
George Orwell

[roquijad]

It's not that easy, F_Smith. When I say "class behavior" I refer to the way a social class sees the world and the way it reacts to it. It's a rather philosophical issue. If you take, for instance, the military class and you want to know what kind of things it wants, you'll find there're several problems. For example, you can find military dictatorships in the world, but they can be fundamentalistic, fascist or communist. MC seems to be obsesed with power in some countries, while they appear more non-political in others. So, for a given civ, what kind of mentality the military class have? Each time you want to add a social class, you have to sit down and think a lot about how people in it should behave and why. We need to decide what the class WANTS to produce values ad hoc for it in the different parts of the model, and that's no easy task. It's not like just adding objects to the model.

[F_Smith]

Rodrigo:

I am sorry if I implied it might be quick and easy to decide on what 'behaviors' to give other social classes. It will take thought, and discussion.

A 'SocialClass' command object has a vector of 'policy' objects (key-value pairs of strings and integers: "Capital Investment" "20"). Which is, incidentally, also how 'Government' and 'Culture' work. And there will be default values for each needed type of 'Policy'. So when making a new SocialClass (or Government or Culture) a person only will have to look at the default policies and then decide how to alter those values.

In this way, if Clash becomes popular, a vast number of Government, Culture and Social Class types will hopefully be made by the legions of players.

[F_Smith]

Rodrigo:

That's fine, for now, I suppose.

But remember, you don't have to define all the social classes' behaviors now. An object model like this is flexible, upgrade-able and scalable. Additional new social classes can and will be up to others, later. What would be wrong with someone deciding to add, say, an 'IntellectualElite' social class? Or a 'MediaElite' one? Since the 'SocialClass' interface they'll implement will tell them all the behaviors they'll need to define, is there any actual extra work to be done to the model itself?

I don't see any . . . it will fit right in.

But for now, we only really need one social class, to test the basic model structure. So no problem-o.

P.S.-- there is actually no reason not to allow all social classes to contribute labor, capital and anything else, at varying levels defined by the 'social class'. The code will loop thru it all on a turn-by-turn basis anyway, and get from each person his contributions (taken from his social class) . . . so it's a quick and simple thing to have the 'LowerClass' (UnderClass, maybe?) contribute, say, x percent of it's total net worth (which could be next to nothing) while the 'upper class' contributes a higher percentage of their net worth (which will be considerably higher, of course). Or you can define the 'LowerClass' as returning 0 when you ask for their 'capital contribution'. But it's up to ya'll.

Just don't avoid this because you think it'll be difficult. It's already in. It's just up to you whether or not you use it.

[roquijad]

I wouldn't be so optimistic about classes behaviors being easy to change, F_Smith, but, well, we'll see...

[Mark_Everson]

I'm at home for a day, and then off to a conference. But I think I can be annoying based on my quick scanning of what has been discussed since I left . Also, I have been working on integrating the government and econ models, and have come to some revelations which I will discuss his briefly as I can manage. I am just kind of going to go through the posts since I left, but this will hopefully be coherent.

Although axi's definitions of the classes are attractive from a fundamental point of view, I don't think they can really work at least as far as the economic model is configured. The basic reason for this is that axi's definition does not allow for poor free farmers (as opposed to serfs). The reason is simple. If the upper class owns all land and capital investment by definition, then farmers near subsistence level will never earn enough to feed themselves and avoid starvation. Essentially, the upper class would always take something like one-third to one half the food that the subsistence farming produces, therefore the peasants would starve. So I think all classes need to be able to own arbitrary fractions of land and capital. This is the actual way that it was at many points in history, so I really objective a whole model being screwed up just so that Axi can have ideological purity in the model. I think it is Much better to have it that way anyway, because you get some realistic game play out of it, like when the Roman Republic free farmers evolved into laborers working for the upper class (or just being unemployed). Now there is fertile ground for social strife...

I think having the middle class is less essential than the point above. But because of its historical significance I think it is worth including. Rodrigo, I like your basic ideas on how to implement a middle class into the model. And as Axi points out, we could indeed make a "synthetic" middle class by adjusting the labor and capital possession of the upper and lower classes. I just think that everyone is so familiar with the notion of a middle class that we might as well put it in, and we will probably later think of uses for it that completely escape us now.

On the coding, F. Smith has a very good point in that the objects in the code need to be flexible. For instance, a scenario designer that wants to do a far-future version of Clash, might want to have a different number of rather different social classes. Ideally we should put all this in his initialization file so that if the scenario designer wants to define new classes with new relationships and new ideologies they can do so.

Axi:
there is some difference in the labor to capital ratios in the different sectors. And certainly the labor to capital ratios will change dramatically with per capita income. I'm not sure if we can include all the factors that you talk about in your most recent post, but I had had similar thoughts.

[Lord God Jinnai]

The ability to have more classes for scerio designers could also be useful if we ever had someone who wanted to create a magical clash world for a later port.

[F_Smith]

Just wanted to verify my understanding here, before I go further with the coding . . . forgive me if this is redundant:

A MapSquare will hold a single Government and a collection of EthnicGroups. And EthnicGroups will contain a collection of 'SocialClasses'. And a 'SocialClass' will contain a population number, a 'Culture', and a collection of policy objects (key-value pairs).

I'll be posting code to this effect tonight, unless ya'll point out any gross misunderstandings I've gotten to.

Then, I think, it's time to put out the prototyper that allows us to build these objects so we can see how easy it will be to customize this game as much as a scenario designer wants . . . you want a 'Wizard' class? Or an 'Empath' class? You can design one yourself!!!