Economy Model Version 0.2

[The Joker]

Why didn't I listen to Amjayee? Why did I have to click submit three times just because I waited 10 minutes without anything happening? Why can't I delete my own post? Why is the lag at Apolyton so heavy?

The truth is out there!
[This message has been edited by The Joker (edited August 28, 2000).]

[The Joker]

Ok, I have done some thinking about the complexity of the model, and right now I actually don't think we need all the complexity in it.

For instance, price elasticity. Of cause food should have a low price elasticity, but all the other goods, couldn't they just have an elasticity of 10%, so people always spend the same amount on a good no matter how expensive it is (the price of the good will just determine how many items of the good is bought). Does it really add anything to the gaming experience if cars have an elasticity of 17%? I don't think so. I think there are other ways where we could turn down the complexity. This will not only help us use less computerpower on economy, it will also (which is the most important thing) let us develop the game significantly faster than otherwise.

What do you think?

------------------
"It is only when we have lost everything
that we are free to do anything."
- Fight Club

[amjayee]

Hello from another town!

I just moved into my new apartment. Currently only my computer is working properly (starting from the basic needs here, too), but tomorrow everything will be settled, and I can get to work again. Until that, I just express my sympathy to Joker. This is a hard task. I agree we perhaps could cut on complexity, but I need to think about it. For me the economics system is the most difficult one; about all other systems I have quite clear vissíon of what thay should be like, but about economy, I quite can't see the whole picture yet. Partly because I know only very little about economics. But in the following days, I hope to get some things cleared out.

[The Joker]

I can not believe how hard it is to create this economy model. It may very well be the most dificult thing I have ever done! At least no schoolwork has ever come close to it.

I agree that we should use Guildmaster's new demand system. But it means that we will have to make the demand function all over again, which means even more work.

What groups should we divide demand into?

Some suggestions:

Food (elementary)
Housing
Luxury consumer goods (like wine and caviar)
Transportation
Communication
Education (primarily for the people's kids)
Longer lasting consumer goods (like clothes, furniture and such)
Entertainment?

What else should be in??

And how excactly should we determine how much money people spend on each type? And a more basic question: Do we really need all this complexity?

I was just thinking out loud here. But I really think it is a tremendous task to make this economy model. I feel that when we have completed the pop, the map and the economy model designing the rest of the game will be a walk in the park. Of cause this is not true, but it makes me keep on going!

------------------
"It is only when we have lost everything
that we are free to do anything."
- Fight Club

[korn469]

i have played a couple of games recently but i think i have came across a gem...i'm unsure if any of you have heard of this game but it's called starknights and it's in beta right now. but the game has two very great features that i think we should seriously look into, one is the economic system. it has one of the most interesting economic systems around. it makes sense and it is complicated yet simple enough to learn in a day. i really love the autotransport system it uses. it reminds me of imerialism....

the turn system it uses is also something that i personally haven't seen before, it is almost like a realtime/turn based hybrid good stuff and definantly worth a look. the download weighs in at 9.8Mbs and even the people with 28.8 modems can download it in a reasonable amount of time...

i encourage everybody to check it out because i think it could help the development of certain aspects of our game...
www.starknights.com

korn469

[The Joker]

Korn:

Unfortunately I have never played Imperialism I or II. But from what I've heard it seems to be a great game with lots of features we could use. Could someone who have played it explain how it works? Or better yet: I will find it somewhere and buy it. It is pretty old, so I should be able to find it for nearly nothing. Besides, I actually miss playing computergames. I haven't really played any since I stopped playing SMAC about 6 months ago.

I will also, as I wrote in the other thread, download Starknights and see what it can offer.


Guildmaster:

I think that we could propably just use 1000 or 10,000 people as the basic pop unit. But I actually think that the demand for food should be far more advanced than you described. I agree that 1 food (or alternatively 10 food) per pop unit should be the basic. This will be the minimum amount. When the amount of food per pop drops below this the mortality of people will rise, and reach a level of 100% when people get no food at all for a turn. But when people become wealthier they will also eat more food. This can easily be seen just by looking around us. We don't just eat grain or potatoes. We eat MEAT! And meat is really not efficient when being produced. If you use the food you use to feat the animals with to eat, it is actually possible to feat 10 times more people, compared to when people eat the meat that comes from the animals. In the western world we have become wealthy enough to demand only the good parts of the animals. This means that when compared to the basic demand for food we eat at least 10 times more. Not in a way that we actually eat 10 times more food, but that the quality of the food we eat is better, which means that it will take more "basic food" to supply us.

This could all be done by having different grades of food, but since it all comes from the same basics I see no need for this. I therefor suggest that when the wealth rises the demand for food will too. But when people eat more than 1 food per pop unit their health will not rise. They will just become wealthier.

I think that the supply function should actually more depend on the money earned per produced food unit. If this amount is high then the supply will rise, if it is low then it will drop. As the demand rises when the price is low and drops when the price is high this will make sure that the supply follows the demand.

I actually think that I have the supply of all the goods pretty much figured out - I have a working system ready (although most of it is just inside my head right now). My main problem and concern at the moment is the demand function. When will people demand what, how will we determine what a rise in income will be used on etc. Especcially with the new demand groups and all (note that I support them, I think it's a great idea, I am just not sure if I understand them).

But you seem to have a pretty good idea on how to work out the demand of people. If you could describe this in details then I am pretty sure I could work out a good supply system to handle all the supply, investments etc in all sectors to work with it.

quote: This would provide the necessary scenarios I think, like the great depression when food was plentiful but no one could afford to buy it so they all starved.

Yeah, this is one of the things that our economy model should be able to work out on it's own, meaning that we should have a realistic way of developing an economic recession. The economy model should have the mechanisms to by it self create a crisis when certain things happend - no special "crisis economy model" should take over just because a war starts.

Another crisis scenario - the oil crisis thing, where increased oil prices means that the costs to buy goods rise, meaning that the producers earn less money on their production, which causes the supply to drop, causing higher prices and therefor lower wealth to people. It also means that the producers will have less profit, causing less investment and therefor higher unemployment.

I think I can work out a way to make this scenario come true.

When we have the model pretty much set we should propably try to effect it in different ways, to see what happends when certain things happend. Thereby we can check if the effects are realistic.

------------------
"It is only when we have lost everything
that we are free to do anything."
- Fight Club

[Guildmaster]

Food (f); D(f)=P, very simple. One pop unit eats one food in one turn. We should use this as the basic number. While it is true that some people eat less and some eat more, generally people across the world given identical financial circumstances and supply would eat right about the same ammount of food on average. If S(f) is less than D(f), people are proportionately unhappy and unhealthy. If S(f) is .5D(f) or less, people starve. If S(f) is greater than D(f) people are happy and healthy and populations grow more. S(f) is primarily a function of farming techniques, rainfall, irrigation, agri-tech, fertility, etc. S(f) is modified by D(f) as follows: if range of S(f) is less than maximum, (ie. if the capability to produce/import more food is available) then the actual supply is the initial supply +10%. Subtract this number from the wealth of each class according to the population of each and the remainder is used to buy other goods. If this is a negative number (the poor can't buy enough food to survive) then S(f)=as much food as the poor can buy until they run out of money. then assume the demand has NOT been met even though the supply is there because the supply is inaccessible. People are unhappy unless S(f) is less than .5 in which they starve.

This would provide the necessary scenarios I think, like the great depression when food was plentiful but no one could afford to buy it so they all starved.

I have to go now, I will do housing later...


------------------
Peace and trust can win the day despite all your losing
-Led Zepplin

[korn469]

ok i will try to describe the starknight economic system but instead of listing the whole thing i will only post a part...it is a space conquest game...in that game you build ships, but instead of being like civ and taking 50 shields to build a unit it takes actual resources

ok for simplicity's sake i am only going to go over part of the system...

lets say we want to build a shuttle (you can equip a shuttle with all kinds of goodies from shields to antimatter missles...it is only a chassis but it will suffice)

in order to build a shuttle we will need the following:

10 ship structures
a space yard
4 turns

ship structures are built out of the following:

3 x advanced chips
3 x aluminum
3 x carbon fiber
a factory
1 turn

advanced chips are made with the following:

10 x silicon (basic resource it is simply mined)
5 x gold (basic resource)
1 x germanium
an electronics plant
2 turns

germanium is produced with the following:

5 x Ge Oxide (basic resource)
5 x energy (resource produced by a solar power plant)
an electrolysis plant
1 turn

the interesting thing about this game is there is a character class profile called production (you choose a character class when you start the game, each has various abilities) if you only have one level in this class then even if you have 50 energy and 50 Ge Oxide you can only produce one unit of germanium a turn. however if you have leveled this ability up enough then if you had 50 energy and 50 Ge Oxide you could build 10 units of germanium in one turn

as you can see i only went down through one branch of this resource tree...but the thing is that to have an advanced economy you will have to have a very large industrial base or some means to buy what you need (you can buy any good you need from the AI or from other players so getting a bad start isn't fatal but it is a setback)

there are a number of planet types each that harvest the same type of resource (but have random amounts) and and each house the same structures (that you must build of course)so if you have a jungle world it will mine Ge Oxide and Al Oxide (stuff to make aluminum) both of which are produced in an electralysis plant

each planet has a demand for certain goods...sand planets which house electronics plants will demand the resources it takes to build advanced chips...now there is one of two ways to get goods from place to place, one is the oldest method in the book...you take your ship load up it's cargo bay with good and ferry them around to where they need to go...but there is also a second method...you have a fleet of freighters which automatically move goods between planets. for each good on a planet you can set a freight attitude for you. you can choose from the following:

store the good on planet
import goods to this planet
export goods from this planet

whats even better you can set the amount of good you want to export to a planet or import from it

the higher the export number is the less the good gets exported from your planet...the lower the level the more good get exported from your planet. it is exactly the opposite for importing goods (high number lots of imports, low number few imports)

freighters are a limited resource that you can build more of...without a number of freighters your empire's war machine will grind to a halt and your economy will suffer, your basic goal is to make sure that goods are moving about your empire, and even more importantly that vital industries get supplied with what they need...the player in effect mostly manages the trade routes

ok now i DO NOT endorse this as our final economic system however i do think that using basically the same system as the bones of our economic system would be perfect

i think that we have to add in some factors to our game to take these bones and put some flesh on them

• *we should incorporate labor into our system
  *each provice should generate demand and it is up to the player to try to set up trade routes to meet those demands
  *the player should not have the ability to manually set up trade routes (except under communism)
  *the player can effect trade routes by various means, some of which could be, signing trade alliances, blockading a trade route, taxing a trade route (can be done at either end of the pipe line), subsidizing a trade route (done frome the export site) but basically the player should be able to reward certian types of behaviors and punish other types, so although the player cannot directly influence the exact outcome, the player can certainly have an indirect influence
  *we should include the quality of goods into our system
  *we should make it so that resource centers need to be profitable, centers that are not profitable should go bankrupt
  *labor should have a cost associated with it
  *transportation should have a cost and a time associated with it
  *transports should also have a need to stay profitable

so what do you think of that?

[korn469]

ok here is a little summary about what i think the strengths and weaknesses of the various systems are...

civ

pros:
-it is very simple, easy for both the player and the AI to use
-it is straightforward, a player spends virtually no time on making sure production lines are running (however much time is spent on maximizing shields)
-it is light on system resources

cons:
-it further reinforces the concept of a civ as a confederation of city states instead of a nation
-with great enough production even the most complicated units can be built almost instantaneously
-trade has almost nothing to do with the actual process of manufactoring
-blockades, trade embargos, worldwide depressions, economic collapse are virtually impossible to simulate
-the system is just simply too basic

starknights

pros:
-it has depth
-provides a basic foundation for something even more elaborate
-could actually behave as an economic model
-takes a number of factors into account that civ has abstracted

cons:
-complicated
-needs to be AI proof
-still lacks many advanced features (labor effect, bankruptcy, and a host of other things)
-needs to be highly intergrated into the rest of the game

that's all for tonight...more to come soon

korn469

[Guildmaster]

It seems this system is more of a supply route thing, we could most certainly use that aspect of it in our game as it seems to work pretty well. As for trade, howevever, I don't see anything here to truely represent it, so there is no real supply or demand other than what you need to build your army.
The main point if I make a point here is that sims have needs too. Obviously, the player creates a demand in that the player seeks out heavy industry capable of sustaining a war machine, so the demand for war machines is high. Sims have needs too, and somehow we need to factor this in our model.
So what do sims want? Well if I were a sim, the first thing I would want is something to eat (I'm hungry) and I suppose I would want pretty much the same thing I want now, a better way to get around, more money, cleaner water, bigger house, less rules, etc.

Ok the food thing:
There are really several levels of demand, first there is just enough food to survive. That could be as little as a bowl of rice each day. Well then there is a balanced diet that is just enough to be healthy, which is perhaps a little more variety, perhaps only 3 bowls of rice per day. And then we figure there is the actual ammt of food people really want to eat. And that, I think, is the real "demand" because beyond that level there won't be much food bought. That's why I modeled my figures as I did.
As for meat, stuff like that, what do you think about this:
If we use graded goods for food, the grade would represent the quality of diet the people are eating...
here is an example of a dinner course for each grade:

Grade A: Caviar, roast duck in menuere sauce served with a light ratatouille preceeded by freshly baked 7-grain bread... Grade A foods purchaced automatically create a +3 demand for C or better wine. Health +3

Grade B: Caesar Salad followed by Mashed potatoes & Gravy side of steamed Broccoli in garlic butter to go with the BBQ Pork Chops. Add french bread as an appetizer. Health +1

Grade C: Choice of fried chicken or catfish, boiled carrots and generic sliced bread. Health +0

Grade D: Burger and Fries no vegetables or side bread. Health -1 for lack of adequate variety

Grade F: Bread and Water. Health -3

I dunno. Of course there would be additional health consequences for eating below the required ammt... so if you're so poor you can only afford to buy half as much grade F food as you need food, not only do you have -3 health for inadequate nutrition, but also you have penalties for not eating enough bread and water.
Ok...
People should automatically buy food before they buy anything else, including housing. However, there are degrees. I do think that a person would be inclined to spend some money on housing even if the minimum food requirement is not met, say if they have just enough money to buy exactly 100% of the food they and their family needs but not a penny more, they might be inclined to spend 70% of this on food and the rest on housing, and thus meet both needs inadequately as opposed to leaving either one unmet. For this reason I think that we should assign a priority to each good in that this is the order they are met...

Ok elaborate:
Each good should have a percantage of the initial demand that MUST be met regardless. Say if there isn't enough money to go around and meet all of them to 100%, it will go down the list and meet each of them in order to the minimum ammt.

Ok example:
Food has priority 1, housing 2, transportation 3.
Food has a starve point 50%, housing has a minimum of 30% and transportation has a minimum of 10%. This means that the sim will try and meet at least 10% of the transportation demand even though it's not eating 100% of the food it needs to survive.
So it will spend it's $100 as follows, spend $50 on food, eating only 50% of what it needs to, then it will spend $30 on housing and another $10 on transportation meeting the bare minimum first. Then with the other $20 the sim will go back and spend it on food because that, again is the first priority, or better perhaps spend a percentage of the remainder.

I have to go now, will post more later

------------------
Peace and trust can win the day despite all your losing
-Led Zepplin

[Guildmaster]

Food (elementary)
Housing
Luxury consumer goods (like wine and caviar)
Transportation
Communication
Education (primarily for the people's kids)
Longer lasting consumer goods (like clothes, furniture and such)
Entertainment?

Ok... I would take these and say my list would be the following:
Food
Housing
Communication
Transportation
Textiles & Manufactured Goods
Labor & Services
Specialty Luxury Goods

Food: Pretty self explanatory, I think.

Housing: From tepees and grass huts to Arcological Complexes, this covers the basic needs for people to have somewhere to live.

Communication: This is a fun one, because the degree of this demand being met can have a profound impact on the game being played in that if you're an ancient empire it could take several turns after you tell a unit to move it actually moves, it slows the information of diseases spreading, etc. It comes in the form of anything from foot messengers without writing (and then some with) to pony express and mobile phones, internet, etc.

Transportation: During ancient times, there isn't too much a need for this as everyone walks everywhere. Perhaps in the upper class there may be a demand for horses and buggys but that's about it. By the time you become industrialized, however, and railroads, cars, eventually planes-trains-and automobiles take rise it becomes a major MAJOR factor. Especially with the rise of larger cities and suburbs.

Textiles & Manufactured Goods: I renamed this from "longer lasting consumer goods" and it includes togas, blue jeans, cabinets, toys, stereos, etc.

Labor & Services: I took education and entertainment and made this category as it also includes medical personell, lawyers, engineers, etc. Basically this "good" is the trade of services by specialized skilled individuals. Remember this is a uber-category and everything underneath it is tracked separately.

Luxury Goods: People who have money like to spend money and what do you give to someone who has everything? Something to spend it on! So you become a wine connoisseur and start eating smoked Gouda and Meunster.

------------------
Peace and trust can win the day despite all your losing
-Led Zepplin

[korn469]

Future Home of Comments
[This message has been edited by korn469 (edited August 30, 2000).]

[The Joker]

Korn:

Check out my economy model. It includes much more than a simple chessboard demo. I don't think creating an economy model like the one you describes is much or a problem. What is a problem, however, is creating a model where goods needs both capital and labour to be created, where we have 5 or 6 different classes, all with different incomes that all interact. Where we have the upper class providing capital and the working class providing labour, where we have a labour market where the cost of the labour effects the cost of the goods produced, which effects the amount of goods people can buy, which effects the cost of labour etc. Where we have to figure out a way to make trade and investment between provinces happend in a realistic way. Where an economic crisis can emerge.

This is why it is so incredibly dificult to create an economy model.

But maybe you are right about something: It might be good to make a simple demo of the economy model, actually program it and see how it works, and then add more depth and complexity to it. This would mean that there would be less blind discussion, where we are not really sure what things will do.

Either way, I agree with Guildmaster. What we need is to make sure that the majority of what happends in economy is not anything you do. It is something the people, as consumers do. Therefor the economy will have to work without player interference. It also means that there will be an enormeous amount of trade routes in the game, far more than it would be good to include on the map. The amount of trade should be so overwhelming that the player could not control it all. In stead the player should set the overall guidelines and make embargoes to his enemies.

quote: 1. people will spend no more than 60 dollars to buy fill demand A

We need to figure out a good way to determine exactly what people buy at certain levels of wealth. Right now it seems like we have each our way to do this.

Allow me to explain mine. It is really simple. It includes a list of goods, which makes it possible for the computer/us to always know what people will spend the next credit they earn on.

Example of a list:

0-7 credits are spent on food
8-9 credits are spent on housing
10-12 credits are spent on food
13-14 credits are spent on housing
15-17 credits are spent on textiles and manufactured goods
18-20 credits are spent on food
etc

These amounts are when the price on 1 item of each good is 10. The advantage of the system is that you can always see what people use their money for. So if some people have an income of 16 they will spend 10 credits on food (which would allow them to buy 1 food - the lowest level where they stay healthy), 4 credits on housing and 1 on textiles and manufactured goods. The list would simply go on and on as people's wealth rises, meaning that even when people earn 2060 credits per turn we will know excactly what they will buy. I am still not completely sure what to do when the price on the good changes. Of cause this will mean that the demand of the good will change too, but the rate at which this happends should be different for each good (the goods would have different elasticity). Due to this it will mean that people will end up spending a different amount of credits on a good when the price on it is 15 compared to when it's 10. And calculating this to determine this exactly is a rather complicated business. And with the graded goods we are adding things really start becoming complex.

quote: Future Home of Comments

???


Guildmaster:

Food:

I think it should be more analogue than that. Even if people only get 40% of optimal level some of them will still survive. If they get 20% some, yet fewer, will survive. I think it would make sence to simply have 1 food per pop unit as the basic amount, where people are healthy, and then if people get 0.9 food the mortality will rise a bit, if they get 0.8 food it will rise some more etc. This will make the most sence. But I must still oppose to having graded food. The main point in having graded goods will be to let smaller civs have the option to specialize in very high grades, and therefor not face too much competition. This makes a lot of sence when it comes to heavy industrial things like cars, but since food is the most basic good to make, and since it does not require any heavy investments to produce having graded food will add nothing but unneeded complexity. I think it is a much better idea to simply have people eat more food when they become wealthier.

Demand generally:

I have described my demand system earlier in this post. See what you think!

quote: Ok elaborate: Each good should have a percantage of the initial demand that MUST be met regardless. Say if there isn't enough money to go around and meet all of them to 100%, it will go down the list and meet each of them in order to the minimum ammt.

This is the problem. There is no such thing as 100% when it comes to the demand of a good. In stead there are real numbers. This is what my demand system does.

What we need to have is a system where we know how much money people got, and how the supply graphs look, and then use this to determine what people buy, how much of it, and at what prices. And this is tricky.

Imagine that 10000 people have an income of 67 credits, the price on food is 9, the price on housing is 13 and the price on clothes is 11. What will people buy? And how much of it? My system, although not perfect, and although it does not include the graded goods or any other flashy tricks, can do this.

Demand groups:
Yeah, they sound reasonable. But I think that any good in a demand group should be somewhat replaceable with another good in the group. This means that people can replace cars with public transportation or wine with caviar (more or less). But you can't replace medical attention with education. I therefor think that there should be more groups in the service sectors. Otherwise the concept of groups would fade away.

------------------
"It is only when we have lost everything
that we are free to do anything."
- Fight Club

[korn469]

guildmaster

what i would love to see in the final version of our game would be this...the player increases military department spending by 300% and decided to order 30 new tanks. the player goes into the unit designer and creates the tank, then various military contractors (controled by the ambient economic model AEM) submit bids on the tanks then the player can decide who to award the contract to...the prices submitted in the bids would be based on supply and demand...but how do we get to that? i think that using a highly modified starknight supply route system could get us there.

to prove my point i think we need a very simple technology test program. so i am challenging the programers to do the following:

The Bread vs Beer Economic Demo

Create a simple playing surface for this test. This will allow a visual reference for the game, and it will also allow some advanced factors such as transport costs to be added into the equation.

An 8x8 square grid would be adequate (same as a chess board).
As for movement rules, lets say that movement can only be up or down and not diagonal and that moving one square cost one dollar.

Set up a number of resource nodes and processing facilities.

Lets have eight grain sites (farms) which each produce 4 units of grain a turn. Grain costs ten dollars per unit (base cost, bidding will increase it).
Then lets have two bread factories which for every unit of grain can make 1 units of bread.
Then lets have two beer factories which for every 2 units of grain they can make 1 unit of beer.
Grain must move from the farms to the factories. For this demonstration it should take one turn for grain to move to the farm, and one turn for the factories to make bread or beer and then one more turn to move the bread or beer to the factory.

Add people to the game (demand).

Add eight squares of people into the game.
Each square will have to eat two units of bread a turn. If a squre doesn't eat two units of bread a turn for five turns it will starve.
Each square will want to eat at least three units of bread a turn.
Every square will want to drink one unit of beer a turn.
If a square eats three units of bread and drinks one unit of beer, then it will next want to eat another unit of bread.
If a square eats four bread and drinks one beer it will want to drink another unit of beer.
So the chart is this:

A. each square needs 2 units of bread
B. if met then one more bread
C. if met then one beer
D. if met then another bread
E. if met then another beer
finished

add in money

Each turn every square makes a set amount of money.
Two squares should make 80 dollars a turn.
Four squares should make 100 dollars a turn.
Two suares should make 120 dollars a turn.

add in costs

the cost of bread is the total transportation cost (one dollar per square moved both from the farm to the factory and the factory to the people)
+
the cost of the grain

add in the rules of supply and demand.

1. people will spend no more than 60 dollars to buy fill demand A, except on turn four before it starves, then it will spend whatever it needs to buy bread.
2.people will spend no more than 80 to fill demand B.
3. people will spend no more than 100 to fill demand C.
4. people will spend no more than 120 to fill demand D.
5. people will spend as much as possible to meet their demands without going over the limits set in rule 1-4.
6. If either bread nor beer is sold in a given turn then it stockpiles at the factory.
7. A factory will spend whatever it needs on grain to be able to produce goods.
8. A factory or a farm will always sell its goods to the highest bidder.

so could any of you programers create this fairly simple abstract of the economic system? then we would have a visual tool to help us iron out the kinks in our system

korn469
[This message has been edited by korn469 (edited August 30, 2000).]

[Guildmaster]

Ok so if I understand you correctly, you are saying that if lss than one food is supplied per population unit, there will be a starvation mortality rate which would reflect the inadequate health of a starving people. I can see that we both agree about this except I'm starting from the top of a demand with a maximum ammt of food anyone would reasonably consume and you're starting from the bottom with the minimum ammt of food everyone needs to consume.
It honestly doesn't matter to me which way we go with this, but I do think you will agree with me when I say that whichever way we go we should follow that same pattern in all trade goods.

Ok so I revise my categories. I didn't want too many trade categories because I was afraid that might bog down the computer but let's see here, exactly how many groups do we want?

So from my previous proposition, remove the service group and add:
Healthcare (incl. both goods & services)
Education (incl. both goods & services)
Skilled Labor
Unskilled Labor

Now in actuality it's quite reasonable to satisfy a need for more doctors by producing more drugs, that way we have what it is in the US where a doctor is nothing more than a prescription-writer. This also goes the other way too. It also includes healthcare costs (something you europeans don't have to worry about) but in game terms even if your gov't pays public healthcare you have to pay for it, while if you don't it will be paid for by the people in one way or another.
Same holds true for education...
But what I'm thinking now that I think about it is this:
I don't think any one thing should be able to satisfy an entire group of goods all by itself. As such, We need to add a consession...
For a group demand, if only one good in the group is supplied, a penalty is added to the supply meaning you have to oversupply to meet the demand. This actually makes sense to me as if there are NO cars in europe at all you would need a whole lot more public transit than is normally reasonable, just like if there were NO public busses in the US the ammt of cars would be even more rediculous than it already is. So say you have the following goods in the transportation group:

Airplanes 0
Trains 0
Cars 3290
Busses 0

You ideally want at least four different goods meeting at least 5% each of the total demand, perhaps you might have 5% of three goods and 85% of another, and you could do that without a penalty but the further you stray from this you incur a larger penalty.

This penalty is negated if there are less than four goods in a group.

Also, say in the healthcare group you have the following goods:
Acupuncturists 10
Physicians 2
Surgeons 23
Drugs 15
Medical Equipment 5

Note that the first three in that group directly affect the employment rate of a province because they are all service personell. They are still treated trade good, because you can import/export doctors by having them train in other countries/provinces much like people in US colleges came to study medicine here with the intention of practicing in their home country. The US is "exporting" those doctors while India or whomever is "importing" them. This is all, of course, in relation to the education demand. Perhaps instead you could treat this as exporting and importing education instead of medical personell. I think it should count as both.

Bye for now more later

------------------
Peace and trust can win the day despite all your losing
-Led Zepplin