Playing the current opendev I get the feeling that the intended cost for infra is the time it takes to build them

Districts cost stability and units cost pop but if you have lost of money you can buy all the infra for each city no matter how many are available

There is a limitation that each one can only be built once (unlike districts) which can also be taken into account

I am not sure how this all is expected to coem together: in my last playthrough I have built practically 0 infra: districts were more then enough

If they had upkeep cost they would definitively not be worth it but as is I could just buy all of them from mid-game onward... not sure what the concept is and how to address its balance

Mausklickmoerder wrote:

Don't underestimate by how much districts increase the industry cost for further improvements. As it stands, you really don't need to nerf districts any further. Though I'd prefer gold upkeep over increased production costs, I don't think such a drastic change will be made before release.

A gold upkeep cost might be better for the current balance of the game than the increased production cost of districts now that I think about it. With how amazingly gold scales into the late game having upkeep costs would keep production more in-line with it. There is an issue that arises from this however and that's by reducing production costs, you also reduce buy-out costs, which may have the adverse effect of making money even stronger than it already is, instead of balancing it. We're already so far into development like you said but it'd be an interesting experiment.

TravlingCanuck wrote:
The great thing is that both approaches seem to work.  I've gone district-heavy and ignored infrastructure, and gone infrastructure-heavy with few districts.  

Maybe that is the goal - I just cant tell even after having played a dozen games now

Sometimes it feels like you can runaway with districts (especially after having built at least one wonder already) but infra can be bought up in a single turn towards the mid game and with no furhter cost to anything

If I would have to hazard a guess I would say that the game might be balanced around a combination of production time and buyout costs

Pop and stability are there yes, but they only seem to be a limitation on player prgoress in the early game

@TravlingCanuck 

Never too late to make changes, just 4 months ago they had that new population growth system, I think they should experiment with district upkeep. Maybe increase the base cost of districts and later era emblematic districts?

Based on the early game play era's we had, I don't see a need to add a maintenance cost to the districts - perhaps that opinion will change for the later era's - but I saw it more as the maintenance was already being included behind the scene.  Many of the districts outputs were pretty low / not very impactful when first constructed but did improve as population and other infrastructure were created to support them.  
In these early era's it makes sense that the population working the district is performing the maintenance and upkeep - and that is reducing the total output that would be otherwise available.