Economic Development Model - Opinions Please?

[Kull]

Just "blue skying", but how about handling slaves as a culture instead of a body of people?

[Kull]

Duh.....which Mark already mentioned!

[Mark_Everson]

Kull:

Thanks for reviewing this, and the comments .

I can't believe I didn't explicitly mention population growth! Clearly this should depend on an overall health level. The overall health level would be derived from:

1) Food Level - Food consumption per capita (effect would increase rapidly for initial amounts beyond subsistence, and then get into a diminishing returns regieme).

2) Health/Water Infra Level - Effective Health/Water infra. I say effective because for a more densely populated area you'd need more H/W infra to reach the same health level.

3) Contraception and other forms of birth control when the people want them (probably determined by culture, education level, and per-capita income)

These factors won't be just added together, we'll have to come up with some simple function to do it, since great healthcare does little good if the people are starving.

On slaves, they are just another flavor of people at the level of this model. The govt and culture models will keep track of slaves explicitly (I think). But in terms of the straight economy they are just a flavor of people who fill the bottom rung of the 'developed sites' scale. This will increase the sharpness of income distribution if we ever get that far...
There will be clear political and social effects of having slaves, but I'm not at the moment handling slaves directly in the econ model. I personally am not certain we need them there. They could always be fitted in if they're really necessary. At this point I'm just trying to keep an already overly-complicated model from getting more so .

[Mark_Everson]

Intentions For My/Don's Proposed Models

My judgement at this point is to stay fundamentally within the model I've proposed for the Econ system. I definitely will add some effects of water availability as included in don Don's model.

In my estimation, putting his and my model together would result in a model that is way too complicated. The models are not Seriously incompatible, although it would take a fair amount of work to merge them. Don's model would treat the demographics and trade within the province in a more detailed fashion than mine. However, intra-provincial trade is already Very innacurately treated to improve gameplay. This is so that we can have decent sized provinces. Clash provinces are much too big for Real economic units in ancient medieval times. Given the fact that a decent model of the world is already badly broken Inside the provinces, its my feeling that adding a Lot of complexity to the system just to improve something else in the general area isn't worth it. Some others also found that model a bit too complicated, which certainly was weighed in my evaluation.

This is not a final decision... We will be closing off the high-level aspects of the models soon, but we could discuss it for few days yet. If anyone would like to cite particular areas of Don's model they think we should use, or contradict my assesment, this is your chance. I firmly believe that open discussion and concensus is the way to go about this.

Don's model has a lot of very useful information in it that I'm currently looking into ways to use... I'd like to thank Don for writing up his model. He's the only one that put forth his own model to demonstrate what Specifically he thought was wrong with mine. I will definitely use his water-availability information in some way. Probably availability of water locally will modify the cost of health/water infrastructure. This will mean that 'dry' areas will have to invest more for cities to grow to a large size.

-Mark

[Adam Smith]

Mark:
Sorry for the delay in getting back to you. Things have been pretty busy here, and I hope my comments will not be too late to be of use.
I read through your model and noticed that it has lots of parallels to both general equilibrium models in economics, and to economic growth models. Here are a few features that I think would help capture the essence of these types of models.

1. Production functions for each province.
A production function is something of the form
CC = K1*(R1^a)*(R2^b)*(R3^c)
Where CC = the total output of the province measured in Clash Cash
R1, R2, R3 are productive inputs
a, b, and c are exponents (want a+b+c<1 so the economy has diminishing returns.)
and K1 is a scale factor.
There are a couple of nice features about production functions of this form. First, a player that has an abundance of one input can, to some degree, substitute for an input he lacks. This is a feature that is largely missing in Civ. Second, the impact of all the technology, productivity, and growth issues can be handled rather simply by this one function. Examples include
A. Impact of technological change:
Technological change can be either factor favoring or factor neutral. A change that would be factor favoring, such as the discovery of an improved mining technique, would increase the coefficient of the factor it favors. A change that is factor neutral (literacy might be an example) would increase all coefficients equally, or, equivalently, increase K1.
Historically, the sum of exponents is pretty low in early years, and increases until the sum of the exponents is close to 1.0 in the modern era.
B. Increased productivity is usually modeled by increasing K1.
C. Economic Growth:
All you need to model economic growth is one or more input that can be stored at a savings rate S (set by the player), and applied to next year's production.

2. A government sector that collects taxes at a rate T (set by the player).

3. A utility function for each province. This helps determine the happiness of the various provinces. The utility function can have a form similar to the production function:
U = K2*(((1-T)*CC/P)^d)*(x^e)*(y^f)*(z^g)
Where P is population and x, y, and z are other factors you might care about such as pollution. You could even make one factor privately provided capital and another factor government provided capital, so that what the government takes away in taxes is given back, but not completely, in public infrastructure. If believe that private investment is more productive than public investment, then make the coefficient on private infrastructure larger than that on public infrastructure. Again, you want d+e+f+g<1. Exponents and the scale factor can reflect various technologies in a manner similar to the production function. For example, the discovery of medicine might increase K2. Note that happiness depends on per capita income. Only (1-T) of total income is available to the populace after taxes, thus decreasing happiness.

4. A simple trading sector.
The necessary condition for trade to be profitable is
PD > PO + C*D
Where PD = price of traded good at destination
PO = price of good at origin
C = transport cost per unit distance
D = distance.
Trade becomes more or less profitable (or not profitable at all) as origin and destination prices change.
Trade also becomes more profitable (or even becomes possible) as transportation costs fall. One effect of technology would be to lower transportation costs (eg., navigation, railroad) making trade either possible or more profitable.

Sorry for the long post. I hope these comments will be of use.

"As an advice giving profession we are in way over our heads"
-Robert Lucas Jr., Nobel Prize-wining economist.

[This message has been edited by Adam Smith (edited June 07, 1999).]

[This message has been edited by Adam Smith (edited June 07, 1999).]

[don Don]

Waaah! I wanna be in the Clash model!
No prob, Mark. I think my ideas need much more work, anyway.

[Dominique]

Excellent work in every respect - my only comment: NO NUMBERS! NONE! NADA! RIEN! NIX!

Please, please, please don't rely on the 1% hardcore gamers who indulge in number orgies - think of the majority who like to understand what's happening by having a superficial glance. PLEASE don't make it too complex - a good game LOOKS like there is much complexity, but the player absolutely mustn't be confronted with it.

No numbers!!!

[Mark_Everson]

Adam Smith:

Thanks for the ideas. I had considered using a production function approach way back when, but I thought it would be a little too abstract. Diminishing returns should come out in the model anyway, but just because of limited sites for production. I'd though of making the diminishing returns aspect stronger than that, but again IMO it was getting too complicated. We will definately have Taxation and trade big time

I do plan on using a utility function for food, goods, and services the people desire, as well as 'infrastructure' they want. This is a little different from the usual approach, but I think it'll work...


don Don: LOL

Dominique: I've put up a poll just to do some research on the numbers/no numbers thing. Lets see what they Really are

-Mark

[Druid2]

Mark:

I guess this is where to ask about the Econ. Model data needs ? The Programming Team needs to know what data you're going to require to be stored for each city or province or area or whatever-it-is-called. I know ~you~ know what I'm talking about, but I'll reiterate for others, so we can continue to get input from all.

Specifically:
what item of data [name or number or etc.]
how big each item might be [length of text or maximum value of a number]

And, if you know, how often do you think you'll need it.
E.g.: "only when a population grows" or "every turn" or "once per game" or whatever.

Everyone expects this may change as we go down the path, but it will be a place to start.

Post it in the Programming Thread

Thanks

-Druid2-
King of the Gypsys
Herder of Cats
Duke of Programming Coordination

[Diodorus Sicilus]

Well, I finished reading the Proposed Economic System, and as I threatened (er, promised) Mark, here's me comments....

The system seems to be based on the idea of Upgrading the economic unit from the City to the Province, a larger, but undefined precisely, geographic entity. Economic Units have varied in size throughout history, and it’s a Good Thing to see an attempt to model this variation in a Civ-type game. The problem may be that the variation is too great to make a good game model.
The reason the city is a focus of CivII/CtP and of historical studies is that the city appears to have been an absolute necessity for civilization to start up. The concentration of surplus allowed Service arena workers to be supported, including those in the religious, political, artistic, and engineering fields that more or less defined the “rise of civilization”.
Which means, simply, that the single city is the economic unit for a lot of the early period: the City State was one of the two “civilized” governmental/political units In The Beginning (4000-2500BC). The other was the God-King Empire in Egypt, and that arose because the Nile River defined boundries and irrigation project size and provided communication and transportation between various sites and cities. In Mesopotamia the individual City State with and without vassels was pretty much it until the 8th century BC when the Assyrians started putting governors into conquered cities and integrating them into their Empire - instead of just burning them to the ground and hauling off everything that didn’t burn.
This indicates the major parameter for your Economic Unit/Province: communication & transportation technology. The Mediterranean became the Cradle of the Roman Empire because bulk goods could be transported all over the area by sea. There was NO WAY to transport bulk goods other than by river boat or ship. Your first Province that spans more than one city will either be elongated up and down a navigable river, or centered on a body of water that allows cities on its coasts to transport and exchange goods and services. Inland, it’s back to City States or individual cities, unless Roman Roads or other Tech advances spread the Province out.
The ship technology allowed sea-based economies enormous reach. English port cities were drawing food from the Grand Banks of Newfoundland, 1500-2000 miles away, as early as the 16th century (first English settlement in New England was not Plymouth Rock, it was a shore base for fishermen in Maine!). Only with the advanced canal building (locks, tow paths, etc) in the 18th century and the railroads in the 19th century did you get a similar spreading out of land-locked, non-river based Economic Units inland.
Your Economic Province Model, then, will have to be based on a Really Good Map Editor. One that plots nice long navigable rivers like the Rhine, Danube, Don-Volga, Tigris-Euphrates, Nile, Hwang-Ho, etc, all of which tied together historical economic entities like your proposed Economic Provinces.

The values you propose for “quality of roads bonus” are off by Orders of Magnitude. To give some Hard Figures, here are Haul weights and distances per day (avg) for your proposed road types:
trail man/pack animal
50-100 lbs for 15 -20 miles
dirt road wagon/cart
500-1000 lbs for 10 miles
paved road wagon/cart
1000-2000 lbs for 15 miles
modern highway truck
80-100,000 lbs for 500 miles
railroad (early) 19th century train
250-500 tons for 200 miles
railroad (modern) diesel unit train
10,000 tons for 500 miles

In addition, remember that the wagon/cart has a Maximum distance it can travel of about 5 - 6 days before the draft animals have eaten the weight of the wagonload in fodder: you either need a large infrastructure of feedlots and spare animals, or that puts an Aboslute Uppoer Limit on the distance you can transport a Bulk Load.

Contest: who can think of a name for Money Unit?
Why not change it with the period?
Ancient: Talent, Drachma, Solidus, Obol, Daric
Medieval/Renaissance: Mark, Ducat, Florin
Early Modern: Pound, Livre
Modern: Dollar, Euro

This would give the Gamer another little indicator of how far his civilization has progressed.

In Development, don’t forget in assessing People’s Desires as a motivator that those Desires change over time due to changing conditions, social situation, or Tech. “Rising Expectations” is one factor in assessing the Need for Services, but another factor is the Social Strata of society, When you have a society including a strata of Arsitocratic Nobility, they will generate Desires for luxury goods far in excess of their numbers, and far in excess of the rest of the (peasant) population.
Individuals or Groups of (Non-Government) people cannot develop any large project by themselves until the mechanisms of financing them are developed: the Corporation or Limited Liability Company, the Stock Exchange/Market, etc. One reason the earliest organized states are called Hydraulic Civilizations is that the only organization with the ability to build and maintain extensive irrigation projects was the government (King, God-King, etc). There was no individual or group of individuals who could handle the finances or organization. Therefore, until the “modern” financial institutions are developed, Development has to be divided into what individuals can afford and accomplish, like building farms, individual bridges, short roads, warehouses, individual piers and trade infrastructure, etc, and the Big Projects that the government Has to handle: canals, entire harbors, long-distance roads, and such.

One historical influence on People’s Desires were Sumptuary Laws which defined the Display you could put into clothes, houses, trappings, etc. These were common in all pre-modern cultures and civilizations, and make a great form of Government break on excess ‘wastage’ of Surpluses on luxuries.

On your model of Consumer Behavior, I suggest that there could be massive changes in the percentages given based on the Tech level of the society/civilization. To an ancient/medieval peasant who never traveled 10 miles from his home all his life, Transportation is likely to be a 0 percentage! Durable Goods, Housing, Clothing percentages will all be very low at the subsistance level, will tend to rise as the average Income rises. I suspect that the original model was based on a modern middle class, not the typical historical model of peasant-noble or subsistance & wealth with nothing in between.

Terrain & Climate Effects:
There is a body of historical opinion (NOT universal, I grant you) that Climactic Change has changed enough to vary the Food Production per site. Rhys Carpenter back in the late 1960s collected data to show that the Dorian Invasions of the Med in 1500-1000 BC and the spread of the Vikings starting in 800 AD were caused by climate changes that made northern homelands less hospitable. The Norse settlements in Greenland were definitely destroyed partially by the fact that the climate there got steadily worse for the century after they settled.

Merchants.
Note that the earliest Trade was by the group: family, tribal, or city organization. Technically, the individual Merchant-Adventurer didn’t arrive on the scene until Coinage (800-500BC) was invented, allowing him to amass a personal, portable fortune. Also note that Bulk Trade only existed by sea. Even apparent Bulk Trade, like the Tin and Copper traded across Europe from the north to the Med, was in small lots: until the invention of the Blast Furnace and Reduction Furnaces in 400-1000AD, nobody anywhere got more than 50-100 lbs in a single ingot of metal: no matter how early you get gunpowder, you ain’t getting cannon without the ability to purify and smelt metal in large lots that those innovations gave! The early metal trade was in small ingots loaded on carts, pack animals, or ships.

[manurein]

Guys, IMO what Diodorus tells here is very interesting.
I think it fits quite well with our models, and we should see which of his ideas (if not all) we could incorporate.
Btw, Diodorus, why dont u join us? Ur historical knowledge seems to be what we need to make Clash as realist as possible (although as fun as possible...)

[Mark_Everson]

Diodorus Sicilus:

Thanks for the thoughtful response. A lot of the points you raise are good. Some of them were already in the system as it exists in my mind, and others would be good additions.

I second manu's thoughts. If you're interested in joining the project, you and your thoughts about the real world and game mechanics would be very welcome . You could participate fruitfully as historical advisor, model tweaker, or wherever you see yourself best contributing. Anyway, if you're interested let me know and we can take this further.

Back to our regularly sponsored thread...

On the scale of economic units being wrong for inland areas in the province model, you are completely right. Square-sized inland provinces is in fact what I started out with for a province model. In that case inland provinces would only slowly grow larger as better bulk transport technology became available. I thought it made for a very hard-to handle system for the player. Another problem was I wanted the political empire sub-divisions and the economic ones to coincide. For these reasons, and others that I don't have time to pursue, I decided to go with something like the present system. Its broken inland for early ages, but gets progressively more realistic as time goes on. If you have suggestions on how to preserve what the model does in a good way (province-level markets etc.), and also fix the problem, I'm all ears. Most of the things I've thought of to fix it either throw waaay too many clock cycles at the problem, or will lead the player into micromanagement by giving smaller provinces an economic advantage.

'quality of roads bonus' Yep, you got me there . What I was trying to get at was both economic and administrative effects for Internal provincial matters. (When merchants carry trade goods they would use more realistic numbers like those you presented.) For the ancient case the bonus would be primarily administrative, whereas in a modern case, more economic. But I probably need to take another look at what those numbers should be. As always suggestions welcome.

The consumer stuff I knew needed modification for non-moderns. The point about luxury needs being much higher with more skewed income distributions is very good. I just need to figure out how to make it work. I think sumptuary laws I would tend to just handle as a tax on clothing (while keeping overall tax burden the same) just to keep the system from getting More complicated.

What the people will build themselves, vs what government is required for is a good point. Though some of the 'people' stuff is also assumed to be government at the level of map squares or villages rather than just groups of individuals. I think we'll need to see in playtesting what feels better to the player. For example if the people never build aqueducts by themselves it gives a clear way that government spending can facilitate economic growth.

On the Merchants, they were meant to be merchant families, or even larger groupings, not single people.

Thanks again for the insightful commentary.

-Mark


[This message has been edited by Mark_Everson (edited June 11, 1999).]

[Foobar]

Don: I found your info about irrigation and irrigation very intresting. You say that the output from farmland is largly dependent on water. What anbout temprature? How big effect does it have? What I had in mind was how big differnce is the food output in e.g. Egypt compared to Britain/Germany due to temprature differnces.

Foobar

[Mark_Everson]

Foobar:

I'm not sure how often don Don checks in here, so you might want to email him.

cya,

Mark

[don Don]

True, since I let myself get snookered into helping with the Civ3 stuff I only come to Clash about once every two weeks. I thought I'd repeat my e to foobar here.

Length of growing season does make a difference, and frost problems. But historically there were two problems in Germany, Britain, Ireland, etc. The ground was too hard to be plowed with oxen and primitive plowshares. The Medieval development of the horse harness and a new design of plow helped greatly. Soils in much of Germany are suitable only for barley without chemical fertilizers, while soils in Britain and Ireland are often too rocky or too clayey for intensive agriculture. Gravity irrigation doesn't work well in hilly terrain without terracing, which is a big investment in land improvement with a small payoff for the types of crops grown.

Egypt benefited from a very predictable flood pattern, which in effect fertilized everything. They need fertilizers today because the dammed Nile (ha ha) no longer floods the fields with silty organics. Mesopotamia got irregular flooding that helped. Most of these regions have easily tilled soils; even the most primitive methods work. The level plains make irrigation easy, and the huge rivers were reliable and accessible.