
I get that a lot. And when it's not a game about elections or campaigns or election campaigns, it's a basically a war game. I like games about politics but that's not what I have in mind when I talk about political games: I think about games where players have power over each other in interesting ways. I have not done a good job explaining what this mean but the contrast with war games - by which I mean games that are primarily about war and not the specific sub-genre that is labeled war games - helps elucidate what is interesting about political games.
Take, for example, Forbidden Stars. A superb game, and one that I really enjoy. When you play Forbidden Stars, you worry about having enough money to buy your troops and whether or not you have the best technologies for the troops you want to buy. You also care about protecting your cities and your positioning around the galaxy - but you really mostly worried about your opponents and how you're going to destroy them. Sure, when you play Forbidden Stars with more than 2 players, you'll have the kind of diplomacy that comes with any multiplayer game - empty promises, deflection, piling on the leader and the like. Though it's more pronounced when you have ways to seriously harm each other, these kinds of dynamics are not unique to war games and exist even in the most non-confrontational games. Even Agricola, when played with people who know the game well enough, will have the same dynamics. If you don't hear the common deflection 'don't go after me! James is in the lead!' in these games it's not because the dynamics aren't there, but because norms of table talk preclude them.
I enjoy the diplomacy and table talk of a war based game, like Forbidden Stars, but I don't consider it a political game. War mechanics, that allow players to fight and directly conquer and/or destroy each other's pieces/areas/resources, are an easy way to raise the stakes and bring out emotional exchanges between people. And when done well, they provide a fertile ground for intrigue. But non-confrontational games have their own ways to create room for intrigue. The most obvious way is introducing lying, bluffing, hidden identities or the possibility of a traitor. All of these are common mechanics that make for good games, but they tend to take over and crowd out other game mechanics because it's hard to care about efficiency in managing resources when you don't really know who's trying to do what in the game.
Let's take this opportunity to think about ways to make interesting, non-confrontational games that are not centered around hidden roles, social deduction or bluffing. How can we make an interesting political game that isn't a war game? To answer this question, let's look at some Euro-style games that get interaction right, and one that doesn’t. Starting with the one that doesn't.

More relevant to our topic, the game suffers from lack of meaningful interactions - in fact, it's probably the most 'multiplayer solitaire' in my collection. There are a few reasons for that, but the main one is this: there is really very little one can do to affect another player. As a consequence, there is little reason to care about what other players are doing. One of the coolest features of the game is that most of it is simultaneous. That means there is very little downtime and turns are very quick. Some turns are good, and you get to do everything you wanted and others are bad and you only do a handful of things. But they all pass quickly, as you execute what you have while everybody else execute their own actions. But it also means that you basically only check where other players are at, if at all, once a turn. And you only do so if you're the kind of advanced player who wants to try and guess what they are planning and which phase they might choose. Now, I say 'advanced' because on your first few games you are likely to focus on your own stuff and not really care about the tiles other people are playing. Even if you're a little experienced and playing competitively, you have to work pretty hard to study other players' tableau and you learn very little: even if you guess what they are after, you don't know what they roll and so you don't really have much to go with when guessing what phase they'll select. The main interaction of the game is the phase selection - a brilliant step - and basically that's the only reason to care about other players. If you're an expert in this game, surely you'll spend a lot of time of thinking about it but for casual players, it's just not worth the effort. The only other interaction is how quickly players are going for victory points and trying to end the game. The result is that in a game that is otherwise really great and well designed, I don't really care who I'm playing it with. It's a great game for introducing the concept of engine-building and luck management, and when I play - I usually talk to my friends about the game as if we just finished playing it. We're not talking to each other through the game, but over its head.
Now switch gears and think about two fantastic games that play around with the worker placement mechanic: Dogs of War and Lancaster. Both of them are, at heart, eurogames that are pretty averse to conflict. In both of them, the winner will be the player that deploys their workers (knights/captains) most efficiently - It's about getting most VPs for your every worker you place, balancing immediate VP gains against investment in infrastructure (your 'engine), and building a path to VP generation that works in the context of other players' strategy. So far so classic worker placement - where the vast majority of interaction happens in figuring out what ways others can block you. Agricola is the poster child of that classic worker placement genre, where players need to figure out what others are doing in order to ensure that no roadblocks will stop their path for victory. I'm not a huge fan of this kind of interaction because it's quite demanding to figure out what other players are doing when you're playing casually, and if you don't try you're just playing solitaire again. Still, there is no denying that in order to play Agricola even moderately well, you have to pay attention to other players' strategies and build an engine that works around theirs, making sure that blocking you would be costly to them.

In Dogs of War, your workers (captains) double as investments chips: they have to be accompanied by 'soldiers' that tilt the scale of a 'battle' between two warring houses. Basically, these so-called battle determine the VP worth of each house shield at the end of the game. It really is more like a stock market than anything else, but the main issue is that you can't help yourself but stepping into the market because every one of your moves is making one stock (house) worth more, and another worth less. In addition, every captain on a losing side of a battle is worth 1 VP for players on the other side so the relative value of each spot is in constant flux, depending on players' actions. And players can collaborate on one battle while fighting each other on another- so there's a ton of room for players to annoy each other while not breaking relationships completely, always having options for collaboration as well.
But really, the best part of Dogs of War is that the most antagonist move - placing a worker on the opposing side of a battle - is also a possible gift to your opponent, because it provides an incentive for other players to jump into that battle on your opponent's side (also making the battle more juicy for that player). Likewise, if you jump on the same side as someone else - helping them win a battle - you're also possibly stealing (or splitting) the prize, and it's likely you're taking the reward they want. So you can't help someone without also hurting them and vice versa - and that opens up space for fun banter, negotiation, back and forth and possible deal-making.
Lancaster is more subtle in some ways - it's a lot more focused on engine building, efficient play and careful planning. At the same time it's much more brutal as it lets you bump away other players' workers (knights) by deploying bigger workers (stronger knights). Bumping someone else lets you do what you want even when someone else needs it as well and it also allows you to mess with their plans. Needless to say, it's a very satisfying action. But it's very costly and you might regret it later, when all your squires are spent and little knights are taking up other good spots uncontested. On the flip side, if you're worried about bumping you can overcommit to ensure you get the spot. Sometimes it'll be worth it but other times - if you failed to read the table - you'll be stuck just like the case where you bumped too aggressively or if you were bumped.
What's great about this mechanic is that gives you a lot more flexibility to plow through the constraints set by other players' actions. It's just another little efficiency game placed on top of the previous one but it lets you raise the stakes - how badly do other people want that spot? Will they be willing to commit and spiral down the sinkhole of bidding and outbidding? Here, you have to guess not just what people want but also how badly they want it. Just like in Dogs of War, this opens up the door for all sorts of fun back and forth - kicking someone out, and then getting kicked out only to come back with a bigger knight and so forth. There's also an arms race element where you really have to consider how important it is to get those bigger knight quickly, by judging how aggressive your friends are about what they want to get from the board.
The lesson is - you can make a great euro-game that has no direct conflict and with great interaction. War games are not necessarily political game and there's no reason economic games can't have great interactions and be political games. Indeed, despite popular beliefs - war is much less political than market. Or, at least, so it seems to me.