Government - Economic models interconnections

[F_Smith]

rodrigo:

I actually completely agree with you (you knew I would sooner or later, eh?) -- 3 classes can be perfectly scalable (i.e. cover 100% of the people regardless of pop, under all circumstances).

As long as we call them 'Economic Classes', and not 'Social Classes', and don't try to mix in social concepts like 'discrimination'.

Altho I have one possible suggestion -- should slaves then have their own economic class?

If so, then 4 classes would scale to include every single economic combination in evey single place and time in the history of the world.

I, personally, support your idea.

[Lord God Jinnai]

Well slaves shouldn't be on the UC/MC/LC...they'd be lower than the LC since the LC has atleast some economic benifits and whatnot while slaves don't. Also the basis for what is a slave is differn't from the LC...i think the term used for sefs is good idea in that a serf was assumed to own himself and thus his life. A slave was not. That is the key differance.

[Mark_Everson]

On economic classes vs. social classes:

The classes in the government model are designed to represent a relatively small number of groups of people that are especially important in the dynamics of governance. Some groups are characterized by their adherence to a particular philosophy -- the religious class is a good example. The idea is that a person in the religious class, whether they are at the top or bottom in the hierarchy, would most importantly prefer the government to consider the teachings of that church, and usually the temporal power of that church, when formulating policy. That is Not to say that we think that in the real world this abstraction is correct, but it is a reasonable enough approximation for a game (I am putting words in Rodrigo's and axi's mouths here...). For those people in the society who aren't particularly involved in some focused effort like administrating government, warfare, or religion, we abstract that the largest differences are along ethnic/cultural and economic lines. And economic model is trying to use some of this information also, leading to some of the discussion we've been having here.

It's clear that the model as it has evolved, greatly simplifies some things. But, IMO, it has enough flavor to give some sort of a reasonable "cartoon" of the factors that really have driven the behavior of governments throughout the ages. And it does so with something like 10 classes participating in the power struggle. With this relatively small number of classes a decent GUI can graphically show the player the dynamics of the internal power structure that results in the final government behavior in a given area. I think the player can absorb, and respond to (and get into) things presented in this way.

In response to the idea of having explicitly different sets of economic, and behavioral classes, I personally think it just complicates things without much return in terms of Fun for the player. So instead of something like 10 classes there might be 30. Instead of just the religious class, now there is everyone from UC religious people, down to enslaved RC people. We would then have to model how the rich people in the religious class behave, as opposed to the poor ones, as opposed to the slaves... I just don't see that gets us very much, and increases the overall complexity of the system quite a bit. Would the more complicated system be more realistic? Probably. But we could make it yet more realistic by modeling individuals in the population... So the question for us as game designers, is which level of abstraction gives the best combination of playability and use of resources (that can be used elsewhere in the game). And I personally feel the overlying different types of classes will actively Confuse the players who are only mildly interested in the government model. I think that's the case because they would need to look at more information to get a gross understanding of why things are doing what they are.

That is my take on this issue... what does everyone think about my perspective on this issue? For those of you who are pushing that we need distinct economic and social class systems, where is the big payoff for the player? I'm sorry, but it just eludes me...

[F_Smith]

Mark:

Actually, breaking the social and economic classes into their seperate functional components simplifies the system, and especially the data storage. That's what lead me to split them up in the first place.

To code the results you're after, you end up with 4 economic classes and 4 beginning social classes:

Economic: UC, MidC, LC and SlaveC

Social: Religous, Military, Govt, PrivateCitizen.

And those 8 classes (with simple code) can produce 64 different types of societies. And, as we said, more can be added. Which gives almost unlimited replayability to the game.

Which is almost certainly going to be the 'bang for the buck', for the player. The most exciting part of this game could be the fact that in different places, in different scenarios, depending on economic and social factors, you can experience and have to deal with entirely different societies.

If we only have 8 or 10 different societies, players will see them all in one, or at most two games. Replayability will suffer greatly, since the game will be the same every time.

The idea that the rich priests of my Theocracy will not act the same as the poor priests of a nomadic tribe I played yesterday is a big plus, gameplay-wise. I, as a game player, will have to handle each differently, if I choose to micromanage to that level.

As for the details on how it'll be coded, do you want them here in this thread? Because it will not be difficult, and actually will be more simple than the other approach. Quick summary: for each 'group' of people, the 'govt they prefer' data will be selected from their 'social class' (religous class, for priests). The 'economic behavior/decisions' of that group, however, will come from their 'economic class'.

So all we have to do is code up the different class objects, and let the game handle their interaction.

Simple.

[Lord God Jinnai]

Well its just my opinion but if we have a religious social class we should have a scientific social class also...now this won't appear in all eras but it will increasinly in many societies take over the role religion section has waned.

Right now even with all those there is no way of showing a society driven solely for the pursuit of knowledge for its own sake. Granted such societies exist only in "golden ages" but there is then no way to show that part of the push for knowledge.

[Mark_Everson]

quote: Originally posted by F_Smith on 07-22-2000 06:01 PM And those 8 classes (with simple code) can produce 64 different types of societies. And, as we said, more can be added. Which gives almost unlimited replayability to the game.

I don't understand your math here... if there are two types (economic and social) each with four groupings, then there are sixteen possible pairwise combinations. UC-military, LC-religious etc. However, these of course aren't complete societies, just classes in your scheme. If there is some deep meaning to the number 64, I don't get it. And what happened to the minorities? Is every single culture going to be broken down into sixteen groups in your proposal?

But anyway, unless you are proposing to have your model mapped onto the behavior of the current government model, and are just leaving the flexibility for a later date, you are proposing a huge change in the whole system. So we should probably discuss this further in the government or social thread.

quote: If we only have 8 or 10 different societies, players will see them all in one, or at most two games. Replayability will suffer greatly, since the game will be the same every time.

Rodrigo's model has something like 8 or 10 Classes. This will in no way represent merely eight or 10 different societies! I simply can't enumerate how many possible societies there are, given that the classes choose from something like 20 different ideologies to get their behaviors. All your expansion of the model does IMO is something like replacing a model with thousands of different societies, by one that has tens of thousands of different possible societies. I firmly believe that the level of flexibility you propose will be completely below the radar of 99% of all players.

quote: As for the details on how it'll be coded, do you want them here in this thread? Because it will not be difficult, and actually will be more simple than the other approach. Quick summary: for each 'group' of people, the 'govt they prefer' data will be selected from their 'social class' (religous class, for priests). The 'economic behavior/decisions' of that group, however, will come from their 'economic class'.

This would completely change the model. At least taking your words at their apparent meaning, the PrivateCitizens would then all want the same kind of government, and only behave differently when acting from an economic perspective. Is that what you mean to say? Or does it appear that way because you are just cutting corners to put it all in a short paragraph? I won't say anything further, because I simply may be misunderstanding what you are proposing.

[F_Smith]

Oops:

You're correct -- sixteen different types of 'social classes', instead of 10. I don't know where I got 64. Which is why we'll still need, in my opinion, the ability to add more.

My assumption is that there is little actual variation in the economic classes' opinion of govt and society -- largely only in terms of redistribution and tax rates. Which I think means will limit the actual playable difference between societies to how the dominant 'Social Class' feels about govt. So that in the old model, all govts will rule over one of 4 types of country -- ones dominated by 'Military' societies, 'Religous' societies, 'Beauracratic' societies, or 'Pure Captialist/Merchant' societies. There will be some variety in the Merchant societies, but it will not have much effect on flavor of gameplay, just on production/consumption values. Is that assumption correct? And by using the proposed model, you add all sorts of other possible govt/societies -- 'Knowledge-based' societies, or 'Ethnically pure' societies, or 'Magic-based' societies, or whatever. Even an 'Anarchy-based' society -- a social class of anarchists! Just anything is possible. The limit is only in the imaginations of future players/designers.

And, most importantly, like all the other data-model changes I've proposed, you don't have to use any of this customizability for the next version of Clash. Your models can go unchanged. I'm currently writing the object builder under the assumption that there will be only the basic social/ethnic classes previously outlined. You'll never know the power is there, unless you want to use it. But in the long run, the replayability can be a major saving grace.

The concept of 'minorities' won't be stored specifically at all. It is a game concept only, not a data model concept. The game data will only store how many Groups there are in a province and which mapsquares they're in. It'll be up to the govt 'controller' code to determine the 'status' of those Groups.

Absolutely, for data storage, each Population 'Group' (EthnicGroup) will have to be subdivided into it's seperate class components. Altho again, this will be transparent to the game player, who will only really deal with them on a 'province' level. But when they take over a mapsquare, all the data will be available to add to the new Province that mapsquare belongs to, and take away from the province that lost the mapsquare.

I'm just providing feedback as to what's possible, what I've found on the battlefield. I'm leaving all the actual model design choices to ya'll. The plan does not have to change just because it can.

P.S. -- the server is back up again. They supposedly were making 'upgrades' that are suppose to speed up the thruput. Notice anything?

I don't.

[Lord God Jinnai]

I still say we should add the knowledge or science social class in. I mean it'll help a lot for the tech model for showing how a society who is ruled by scientific elite will pull toward "pure" scientific discovery for the sake of discovery while still advancing others at a better rate. I don't see how outof any combination of the 8 classes you have how you could come up with something like that.

[Mark_Everson]

F_Smith:

Okay, if it's all hidden "under the hood" of the object model then I have no gripes... perhaps you'll convince us in the end! I just keep taking your comments on how you choose to implement the object model to imply you think that the game model must be like that. I'll try not to make that mistake again...

LGJ:

Ummm, other than science-fiction I can't think of a society like the one you describe. That's one of the reasons why we will have customizable social classes for scenarios. But anyway, your comments should really go in the social model thread, not this one. The purpose of this one is to look at the intersection between the economic and government models.

[axi]

I'm back from my trip and ready for action! Yesterday I was busy reading the posts and compiling a (long) reply. But I'm afraid I will have to go away soon. Clash may be a good reason for me to stay in Athens, but it isn't enough this time of the year.

One general remark: Are we discussing about the implementation of the interconnections mechanism I presented in the spreadsheet or are we making a general discussion about government and economy interconnections? Are we doing the latter from ground zero or based on what I have done so far?

1) PP and infrastructure: I don't think that labor can ever be considered as an item of property; it is a factor of production generated by the working classes every turn and used up along with resources. One can say that infrastructure can be an item of property which, unlike kapital, has no product yield (a workshop can produce wealth while a house cannot; hence the first is kapital and the second is infrastructure). But this means that we consider wealth in it's narrow material sense, which is not the case with Clash, otherwise what do services stand for in the econ model? A house can be kapital in Clash because it provides housing services (either free if it is private, or at a price if it is on rent or part of a hotel enterprise). Then almost everything is considered kapital in Clash, because everything (including houses, cars, computers, clothes, furniture and other consumer goods) could be leased for a price? Of course not; but then what distinguishes a factory from a residential building and (say) a shipload of rice? Well, we can say that the factory is useless for anything else than production of wealth, as well as the shipload of rice cannot be used for nothing other than consumption, while a residential building can either be "consumed" (lived-in) or be used to provide services (as a hotel). Now, in modern society property rights are legally recognised to all of the above three types of items, but which should PP correspond to?

I think it should be addressed to all items that ARE used for the production of wealth (that means all kapital goods and part of the infrastructure). Consequently, all other items of infrastructure have to be consumed by their owners alone and investment in infrastructure is regulated by the personal need for the certain type of infrastructure and not by the need to increase the income (to invest).

This means that, at this point, it doesn't matter who is the actual proprietor of this infrastructure, since there are no profits from it. It would only matter if we were monitoring the exact wealth of individuals instead of the wealth of economic classes and were able to cover all the transactions between them and the state (remember my approach of "winners" and "losers" and the way one class's investment is granted to another along with population shifts); but that is impossible at such a scale. That is also the reason why we shouldn't worry much about the nationalisation - privatisation issue. So far, the economic classes don't invest according to the expected returns, but according to their income so such things as profit margins would not interest them (Well, maybe if we are going to include foreign investment, profit margins will have to be included, to redirect investment from country to country, but that's way ahead). So things can be kept simple.

Because almost all of the infrastructure can be turned into kapital, it is possible for the people to reduce their (proportionate or absolute) consumption of a certain infrastructure type, in order to initiate an enterprise (As we Greeks have done with housing to initiate tourism or as the English have done with education to attract foreign students). It is intersting to notice that the capitalisation of infrastructure (as well as the use of capital for the production of consumer goods) has no practical macroeconomic effect if it concerns the internal market of a civ or a province, but (as with my two examples), it is important for the external trade (If there is cheap access to infrastructure provided elsewhere, there is no need to invest in it locally; in the same way as if cheaper goods can be procured from abroad, there is no need to produce them locally).

Where I think we can and should elaborate further is the sectors which the state controls. An economy with nationalised industry is quite different than one with collectivised farms or one with public finance, education, health, transportation, etc... If we think that this distinction matters in any way, we have to find a way to include it.

Maybe my reasoning is somewhat abstract and chaotic, but I hope it will help dissolve the doubts you have about PP. I think all is OK and no changes are needed here.

2) Tax Rate: Rodrigo wrote:

quote: If so, a person gets a wage from the State. Now the State ask him to give money back? Why not just pay him less? I really don't see the sense of using tax rate in this situation.

Now this is exactly what happens to my parents who are civil servants; they pay taxes exactly like everybody else; the only difference is that they get deducted from their wages in advance instead of being paid back afterwards. In my spreadsheet you will see that I apply taxes on everything, even on public profits. OK, I know I have altered the definition Rodrigo gave for TR in the govt model, but this logic makes sense for better bookkeeping and it's nothing to worry about - only the effects should matter. We should not forget about centralisation too; in a federal state without PP (feudal or socialist), taxing the state profits makes sense, so that the local authorities pay their due to the capital.

3) EP: Public ownership doesn't ness mean state control. A anarchist state f.e., would have common ownership but the control would be decentralised into local worker's and community councils. An economy with low EP will be generally left to function without intervention from the player, regardless if it is capitalist or socialist. EP is the most independant of all the policies and can be anything, regardless of SP and PP. As for the question if it can be replaced by Ruler.polpower, EP represents the extent of the executive power over a specific aspect of society (economy) while R.pp represents the player's share of the legislative power. EP is needed for maximum control over the economy without the need of so big control of the political scene. For the time being the use of EP (see the spreadsheet) is not compulsory (It defines a maximum range of player intervention, but doesn't oblige the player to use it). If the team feels this is not OK (Since economic planning means that the state makes all the decisions, even if these decisions are the same as the will of the people, the very fact that the state is responsible for them makes the situation quite different), we could give EP an effect more like the one SP has.

4) Minorities: Although I aknowledge that having minority populations to be other than "slaves" is historically more accurate, I find it quite complicated. Since discrimination is based on religion as well as on nationality, then, if both RD and ED are other than 0, then apart from the majorities there will be, civwide, at most 4 other sets of people which will be discriminated at different degrees each. It would be impossible to manage them the same way we do with majorities. I propose that the interconnections model only keeps track of the percentage of these minority groups that is a victim of exploitation, while all the others are considered economically (but not politically) equal to the majorities population and are included into the majorities part of the calculations. All people in the Exploited Class (XC, or the "Underclass", we must find a name), regardless of their EG are under the same level of exploitation, exactly like Minorities were treated until now. The formula for the percentage could be something like:
EMP=1-(1-RD/5)*(1-ED/10), where RD and ED are considered 0 if the EG has the DR or the civ's nationality.
As for the PAFs, IMHO each EG will have an NRF, in which all members of the EG will participate, while the XC of each EG will have an ADF (noone else will) and also the whole XC will have a civwide RF.

This approach maintains the simplistic admittance that all EGs have equal demographics (Which Rodrigo has made and I had fruitlessly questioned a long time ago), but only for those that are "free". Notice that even if SL is 0, some economic exploitation persists (see the spreadsheet). You see, even in our age of freedom, immigrants are victims of quite heavy exploitation. (Mexican workers without the green card in the US work in sweatshops; Mexican millionaires legally resident in the US invest in the Wall Street; neither of them can vote). What do you think?

My only concern is the lack of a selection mechanism for the discriminated EGs (like the long gone Racial Discrimination, which was a range for acceptable racial distance). Right now it is "in or out", for simplicity, but that is too rigid. All religions are not the same: there are other GRW branches, other GRWs, PRs; also all nationalities are not the same: there are barbaric cultures or civs, friendly or hostile, etc. I know that this belongs purely to the govt model but since it's related, I ask this here. If there is an answer already somewhere, please spare me the tomatoes.

5) Middle Class: It's easy to say "we want the Middle Class in", but it isn't enough. Are you content with my implementation? Rodrigo wrote:

quote: BTW, I like Axi's definition for MC: "a class that has enough kapital to employ their own labor".

But how about it's labor? How much is it? If it's the same or alike as the LC labor, then the ID between these two will depend solely on the current LCR. If it's price is augmented acording to some education level (to represent the concept of the skilled labor as the origin for the rise of the Middle Class), the MC will be even richer and will hold more of the private capital. Will this mechanism ensure the historical rise of the MC (given of course a high enough PP)? How sensitive will this system be to economic changes? I repeat: IMHO, for this and other reasons all econ changes must not be abrupt. If we do not scrutinise this stuff now, I'm afraid we might do something terribly wrong.

6) Economic Classes vs Social Classes: I too agree that Economic Classes are different than Social Classes; in fact I think I was the first to grasp their difference since I introduced the economic role as the distinguishing factor between the UC and the LC. But this doesn't mean that I agree with F_Smith's suggestion that economic and social behavior are two independent dimensions. I can't visualise a society possesing 4*4=16 classes; my idea is rather of 4+3=7 classes. "Economic Class" and "Behavioral Class" will inherit their characteristics from the generic "Social Class" and have in addition economic or behavioral data, but not both. Of course there are always rich and poor, upper rank and lower rank among the priesthood, the military and the state apparatus (although things like slave bureaucrats or slave warriors do not make alot of sense), but these are not the result of economic role but of a hierarchy. Poor priests do not essentially do something different than rich priests, they all provide only Ethics; neither Labor, nor Kapital. Consequently their income can only derive by the value the specific contribution (f.e. Ethics) is given by the society. That's why in a Theocracy all priests are rather well off, regardless of rank. Mark wrote:

quote: The idea is that a person in the religious class, whether they are at the top or bottom in the hierarchy, would most importantly prefer the government to consider the teachings of that church, and usually the temporal power of that church, when formulating policy. That is Not to say that we think that in the real world this abstraction is correct, but it is a reasonable enough approximation for a game (I am putting words in Rodrigo's and axi's mouths here...). For those people in the society who aren't particularly involved in some focused effort like administrating government, warfare, or religion, we abstract that the largest differences are along ethnic/cultural and economic lines.

And he was right to put words in my mouth; because of difference in the RC's contribution to society, the class cannot be divided along economic role lines and so it cannot be divided politically. IMHO, behavioral classes should be totally abstracted and neither their population nor their income should be tracked. Only their preferences, their PAFs and their pol.power (as the legal or de facto influence of the religious/military/bureaucratic apparatus on society) should be taken into account. This is an abstraction, like when we divide the whole income range into economic classes.

Our only preoccupation should be about the pol.power: do the Economic and the Behavioral Clases share the same political power field? If not, does this mean that the same people can be the UC in one field and the RC in the other? What about complex classes with multiple contributions? And if we do not like the abstraction and we want to count priests, taxmen, scientists and elvii, then we'd better find a way to evaluate their contributions to society and add them into the market system. THAT would be novel!

As for Lordy's petition to include scientists, I'm for it. A scientific elite isn't ness a sci-fi theme. There have been various civilisations where knowledge was important and the sages got to form a class of their own, quite distinct from the religious one. The ancient philosophers, the medieval alchemists, the rennaisance artists, the encyclopedists of the enlightenment, the various sages and intellectuals of all ages, have been quite distinct as classes. They would be an easy choice to back up in order to speed up the discovery rate.

Note: Umm, Rodrigo I think you forgot to iclude the population increase in the turn macro you made for the spreadsheet. You might want to fix this sometime...

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

[axi]

double post!
[This message has been edited by axi (edited July 25, 2000).]

[axi]

triple post!
[This message has been edited by axi (edited July 25, 2000).]

[F_Smith]

Hi, Axi:

In the current model, actually there will only be 4 different 'class' behaviors (social classes, that is) -- RC, MC, GC and 'private citizens' (everyone else). Because you can't define 'class' behaviors along economic lines. For example, never do all the 'upper class' of any society share any given 'social' behavior charateristics. I.E., 'New money' v. 'old money'. While their 'economic' behavior may be similar, their 'social' behavior is quite different. Ditto for the 'lower class'. I absolutely have to store seperate 'economic' and 'social' behaviors for each group of people. And for my data model, I didn't see any justification for having an economic class inherit from 'social class' -- it made more sense for the 'group' to encapsulate (contain) info on both, instead.

But again, ya'll will not notice any difference unless you want to create more complex population behavior.

Personally, since there will only be 4 (or 5, if we include a 'scientific' social class) different types of social behavior in the current model, I personally think that this type of open-ended customization will be absolutely necessary to create the depth that will allow people to stay interested after playing the game for the 100th time.

Different 'social' classes of priests, bureaucrats, warriors, etc., absolutely has happened many times. Many societies have had slave administrators. It was very common in ancient Rome to have 'slave' bureaucrats working for the 'upper class' bureaucrats, and 'slave' warriors have been a very important source of supply for many navies. And unless I mis-remember my history, weren't the Persian 'Immortals' who made the mistake of trying to invade your Greece oh so long ago slaves?

[F_Smith]

To All:

I'm about to post what I'm sure will be another unpopular opinion, so please forgive me in advance:

This idea of 'testing' your ideas in a 'spreadsheet' is probably not good.

The spreadsheets do not have to deal with gamedata, as it exists in the game world. This is a very important thing. This has lead to making some design choices that may have been best for the spreadsheet, but did not turn out to reflect actual game conditions (and were certainly not the best choices for the game world).

It also has lead to model integration problems, once the game world is taken into account (hence this thread).

Serious assumptions have also cause serious misunderstandings. The idea that pop and econ data would not be stored at the mapsquare level (a game impossibility) works just fine in the spreadsheets. But it is highly inaccurate for the game world. So any decisions taken on the basis of that assumption may very well have been wrong.

And finally, and most importantly, these models in the game world will be defined by how they work over time. In development terms, that means that the model hasn't even begun to be tested until you are able to run it thru a few hundred years. All ya'll are doing with the spreadsheet now is defining objects (and defining them in an isolated world, at that).

This would be better done in code.

Ya'll absolutely should switch to coding up gameworld 'builder' tools to test your models. Otherwise, from a 'project management' standpoint, this is unproductive time. Theorizing with no testing.

I'm sure Rodrigo would agree that all theories need regular testing.

The scientific method is, after all, theorize, test, revise, repeat.

Ya'll haven't moved on to the testing stage yet. It's time.

Again, I'm sorry if this is going to upset anyone. But your game can/should be under development. And I'd like to help you with that.

Sorry. And feel free to flame me to high heaven.

[F_Smith]

To All:

I'm about to post what I'm sure will be another unpopular opinion, so please forgive me in advance:

This idea of 'testing' your ideas in a 'spreadsheet' is probably not good.

The spreadsheets do not have to deal with gamedata, as it exists in the game world. This is a very important thing. This has lead to making some design choices that may have been best for the spreadsheet, but did not turn out to reflect actual game conditions (and were certainly not the best choices for the game world).

It also has lead to model integration problems, once the game world is taken into account (hence this thread).

Serious assumptions have also cause serious misunderstandings. The idea that pop and econ data would not be stored at the mapsquare level (a game impossibility) works just fine in the spreadsheets. But it is highly inaccurate for the game world. So any decisions taken on the basis of that assumption may very well have been wrong.

And finally, and most importantly, these models in the game world will be defined by how they work over time. In development terms, that means that the model hasn't even begun to be tested until you are able to run it thru a few hundred years. All ya'll are doing with the spreadsheet now is defining objects (and defining them in an isolated world, at that).

This would be better done in code.

Ya'll absolutely should switch to coding up gameworld 'builder' tools to test your models. Otherwise, from a 'project management' standpoint, this is unproductive time. Theorizing with no testing.

I'm sure Rodrigo would agree that all theories need regular testing.

The scientific method is, after all, theorize, test, revise, repeat.

Ya'll haven't moved on to the testing stage yet. It's time.

Again, I'm sorry if this is going to upset anyone. But your game can/should be under development. And I'd like to help you with that.

Sorry. And feel free to flame me to high heaven.