It makes sense to apply "economics light" to most games if for no other reason than to avoid really obvious tropes.

An example of this to have inns and merchants reflect the area they are in. Many RPGs just have the merchant dole out better and better equipment as the game goes on with no regard for logic, but even in a very linear game it can make sense to have merchants sell to their strengths. Smaller, rural towns could have an awesome potion dealer, whereas that mining town could have great weapons. The towns offer better equipment than before, but it's not just the predicable cycle of "enter new town, buy better equipment, move on". I find basing your decisions on something other than RPG tropes helps you to look at the game and say "wait, why am I doing this?" which usually ends up refining the game making progress and giving you interesting ideas on how to overcome challenges.

Setting up regional economies works better in big, open-ended games where there is a tonne of backtracking, like Shinan said. It's tricky to balance fun against micro-management of items, though.

Basing it off of locale in a linear game would be pretty interesting.

Imagine that you are in a desert and potions are the only item that heal you in a game (or at least currently). These potions happen to use water as a main component to them. Since water is scarce in the desert, it drives potions up to a ridiculous price. Perhaps reaching over 3000% of its typical value. Now, if a player was in the desert, and had no way to access other shops, they might be inclined to purchase some of these exorbitantly priced potions.

If you were able to afford the player enough income to purchase them, and presented them with a situation where they were absolutely necessary, then it would definitely add some interest to the game. That player, when they first see the prices will be like "Man! That's way too expensive, not buying those..." and then later "I can't believe I'm actually buying these."

As far as fluctuating prices go, Chainer's got the right idea that they should be story-driven. It'd be cool to put a real-time economy into a game, but I am pretty sure that the actual math behind it being credible is beyond the intellect and/or effort that the designer can put forth. In my experience, even the pro games that try to do this end up with something akin to a number game. Unless the math behind this is solidly backed and dynamic, there's pretty much no way that an artificial economy won't become this. Once a player learns of any piece of the algorithm behind it, they can slowly deconstruct the system until they are able to manipulate each piece. Basically, a real-time economy would have to be encrypted through all of the minute details and random bits that give it real-world relevance.

With all that being said, I've considered the very system I just laid out on occasion. The conclusion I've come to, is that it would be better to let the player indirectly affect market prices, and leave the random and small-scale fluctuation to the game itself. In this sense, the player might drive up the prices of anything constructed with ore in an area relative to the mine he just collapsed while battling a giant dwarf. Buying out a shop of its potions would just result in having to wait as they restock the shop. Do it enough times and the shop might restock faster, as the owner tries to keep up with demand. Do it enough times after that and the owner might stock new types of items with the extra surplus he's getting from you buying his potions. In this sense, it is possible that the prices of potions eventually rise as the producers of the potions start running low on the material components of the item, or the owner seeks new potion providers who might charge higher prices because it is less convenient to ship to him.

In this sense, the player might drive up the prices of anything constructed with ore in an area relative to the mine he just collapsed while battling a giant dwarf.

I was thinking about this, as well, as it ties into a greater philosophy I have about having the player indirectly affect the game world to improve design aspects as well as logical ones. The player is going to go and kill Blaggi the Giant Cave Dwarf because it's a quest and the player loves quests. Well, after he kills him, he's going to have leveled up from fighting Blaggi's minions, part of the Cave Dwarf Miners Union Local 26, so he'll need better items. Luckily, destroying Blaggi and his dirty union henchmen has allowed the city's honest non-union miners access to excellent iron ore, so in they go. The smiths sell better swords, now! This lets you reuse towns, since now the player is setting off in a different direction from the same point. This allows designers to focus on making a couple good towns rather than a plethora of totally interchangeable ones. Similar ideas are players opening merchant passes, shipping routes, et cetera, et cetera, to change the challenge of any given area rather than have each place he visits contain precisely one hardship. One of the reasons I like thinking about logical things and encourage others to do it isn't because I think realism is a must in every video game but because it forces your brain to think up interesting solutions to challenges you've created for yourself, rather than just write them off as being "just a game". The player usually appreciates the extra effort.

Buying out a shop of its potions would just result in having to wait as they restock the shop. Do it enough times and the shop might restock faster, as the owner tries to keep up with demand. Do it enough times after that and the owner might stock new types of items with the extra surplus he's getting from you buying his potions. In this sense, it is possible that the prices of potions eventually rise as the producers of the potions start running low on the material components of the item, or the owner seeks new potion providers who might charge higher prices because it is less convenient to ship to him.

This COULD be fun, but it sounds a bit like an expansion of what Morrowind did, and that was no fun at all. Each merchant had a set amount of gold each day, which you obviously wanted to get since it weighed nothing. Except some of the items sold for 60000 gold and the most merchants carried was 7000, or something. Which meant if you wanted it, you had to trade for their ENTIRE INVENTORY (in barter), then rest for 24 hours, and sell items piece by piece over the next couple days until you got it all and shrugged the heft of those items off. I'm not saying it wouldn't be easy to make a better system than that but the trick is in balancing it. You don't want the player to get swamped in an area because they are trying to level up to get past the next area, but need loads of potions to keep going, but you also don't want to make the system's effects so negligible that there is no point in implementing it, either. I mean, what if I really need thirty potions but the merchant only has ten? I'd have to wait around for him to get more.

This system would, however, lend itself well to a game that had lots of merchants all vying for their market share; one potion merchant sells out, and the next day all the other merchants raise their prices for health potions. This happened a lot in medieval times in local economies. It could also be interesting to see implemented in a sort of Uncharted Waters that was set entirely in one huge city where you are trying to make money as an adventurer. You've settled down, and are trying to run the best business in town by controlling the town's economy, all the while making gold slaying orcs for the city guard on the side. Hmm, could be interesting!

I have been tempted by the thought of an RPG with something like a stock market you can participate in, or like the auction house in WoW where you can manipulate the market for profit.  But imo, this would just be a silly, not especially interesting mini-game.

My idea of economics in RPG's, though, is to just have differently costed items in each town.  For example, a town near a silver mine would have cheaper silver weapons than other towns, towns with some sort of Harry Porter wizardry school might have cheaper magical items like +1 Swords, and so on.

I was thinking about this, as well, as it ties into a greater philosophy I have about having the player indirectly affect the game world to improve design aspects as well as logical ones. The player is going to go and kill Blaggi the Giant Cave Dwarf because it's a quest and the player loves quests. Well, after he kills him, he's going to have leveled up from fighting Blaggi's minions, part of the Cave Dwarf Miners Union Local 26, so he'll need better items. Luckily, destroying Blaggi and his dirty union henchmen has allowed the city's honest non-union miners access to excellent iron ore, so in they go. The smiths sell better swords, now! This lets you reuse towns, since now the player is setting off in a different direction from the same point. This allows designers to focus on making a couple good towns rather than a plethora of totally interchangeable ones. Similar ideas are players opening merchant passes, shipping routes, et cetera, et cetera, to change the challenge of any given area rather than have each place he visits contain precisely one hardship. One of the reasons I like thinking about logical things and encourage others to do it isn't because I think realism is a must in every video game but because it forces your brain to think up interesting solutions to challenges you've created for yourself, rather than just write them off as being "just a game". The player usually appreciates the extra effort.

That would be a great way to expand on that scenario. One of the cool things, which you've basically stated, is that by thinking logically, the game will naturally develop its own architecture around the story. I think this delves more into story aspects than economic ones, like what sort of impact do economics have on one's story? What's interesting, is the sort of effect economics can have on a social situation in the real world, and how that can transgress into a game's story when interpreted from real-world applications. Such as the mining situation we've laid out.

This COULD be fun, but it sounds a bit like an expansion of what Morrowind did, and that was no fun at all. Each merchant had a set amount of gold each day, which you obviously wanted to get since it weighed nothing. Except some of the items sold for 60000 gold and the most merchants carried was 7000, or something. Which meant if you wanted it, you had to trade for their ENTIRE INVENTORY (in barter), then rest for 24 hours, and sell items piece by piece over the next couple days until you got it all and shrugged the heft of those items off. I'm not saying it wouldn't be easy to make a better system than that but the trick is in balancing it. You don't want the player to get swamped in an area because they are trying to level up to get past the next area, but need loads of potions to keep going, but you also don't want to make the system's effects so negligible that there is no point in implementing it, either. I mean, what if I really need thirty potions but the merchant only has ten? I'd have to wait around for him to get more.

This system would, however, lend itself well to a game that had lots of merchants all vying for their market share; one potion merchant sells out, and the next day all the other merchants raise their prices for health potions. This happened a lot in medieval times in local economies. It could also be interesting to see implemented in a sort of Uncharted Waters that was set entirely in one huge city where you are trying to make money as an adventurer. You've settled down, and are trying to run the best business in town by controlling the town's economy, all the while making gold slaying orcs for the city guard on the side. Hmm, could be interesting!

Yeah, Morrowind was pretty terrible in that regard. Even Oblivion maintained the same quirks iirc. But, that is specifically what I'm talking about when a game half-asses something like economics. You end up with some sort of number game that will eat up the player's time while they try to exploit it. In this case, it is straight-forward; you need to pump the merchant full of gold in order to get a greater amount of gold in return. The problem with Morrowind, was that there was no real place to sell all this amazing stuff you've found. As, all of the merchants were poor as dirt in comparison to your character in later levels. Introducing some characters who were interested in buying these items, and could afford them, would have solved this issue and could have potentially developed plot arcs as well. But, then you get the question of how much time can be devoted to such and such aspect. And, basically, introducing something like that would have eaten up valuable resources that could be spent elsewhere.

As far as the potion scenario entering that same realm, it is possible to create a situation where the likelihood of such would be improbable. One solution, would be to address the notion that potions are cheap and dispensable by increasing their effectiveness and their cost, and presenting the player with combat/adventuring situations where using them becomes strategic rather than simply life-sustaining. Under this plan, it would be possible for a merchant to better react to his dwindling potion supply, simply because the effects of which from the player are not as immediately detrimental to his stock. This is under the assumption that the player adheres to the artificial guidelines of what a potion is worth, however. In order to accurately present and impose this, would take skill on the part of the game designer. What this brings up, though, is a completely different game than your typical jrpg. I think, the deeper you go into the realm of presenting economic situations to the player, and balancing them out to prevent exploitations/nuissances,  the further you're going to get from a leisurely rpg. This is pretty well represented in the divide of rpgs that offer some rudimentary form of economics and those that don't.

I have to say some of these ideas sound interesting. The closest I'd gotten to anything like this in a game idea was pretty far from the full economics you're speaking of, but still a subversion of the normal fantasy tropes. In the setting, potions are made one of two ways- mass-produced in factories (Yes, I said POTION FACTORIES) that simply channel magic  from a crystal into prepared bottles of solution that slide by on a conveyor belt, or the old-fashioned way, by hand with lots of natural ingredients- artisanal, if you will.  The artisanal ones, are more expensive- not only because they're made in small batches with high-quality ingredients, but because they tend to give back more HP than their counterparts.. In fact, one of the party characters is studying under a healer to prepare her for eventually taking over her father's  potion-making  corporation- she hopes to raise their factory-made products to a somewhat higher standard.
And - a digression- a lot of the cheap potions are artificially flavored (Grape, strawberry, melon, etc) and some characters will either refuse to drink ones they dislike or get back less HP from them.  The witch is weak and you've got only artificial grape potion, which she despises- you're screwed.