Thank you for your articles, very well done. Re this questions (why larger bldgs aren't much cheaper...), as you probably know, the rule of thumb (for most buildings) is that the bare structure (which is much of what you focus on) is only 20% of the cost of the finished building construction cost (ie, omitting land cost). The other 80% are interior finishes (finished floor, walls, fenestration...), MEP etc. These 80% scale almost linearly with floor area. On the other hand, I agree that the economies of scale not occurring is very puzzling - I'm still astounded to go by US sites and see stick-built construction still happening, even on 4 or 6 story apartments over concrete podiums. Its real piece-work. In northern Europe there are many more economies (eg, panelized construction - they tend to replace more labor with capital). Part of the reason is that, surprising to me, 2x4 wood framing is still the cheapest option in N. America for small-medium residential construction. Again, thanks for thoughtful pieces. (PS - would Munger live in that dorm? doubt it. the psych costs would appear terrible).
As Charles points out, most building costs are interior fit-out, and that's even more true for single-family homes where the structural system and design costs are a pittance. One of the greatest challenges for prefabrication methods is that they stall out at the structural level, maybe with some architectural finish. We don't do much prefabrication for either MEP or interior finish work. That's true of precast concrete and tilt-up systems, panelized wood structures, both cold-rolled and light-gauge steel assemblies, glazing modules...
What I'm not really sure of is whether we *can* do that economically; talking to the professionals, it seems no one is trying.
For example, is there an impediment (genuine or code-based) to creating a fully panelized, factory-assembled system for single-family home walls, where utilities "plug into" standardized connections in a wider building system? For single-family homes, such a system would seem to be within reach if I'm building several hundred similar family homes in 8 sub-divisions in a metropolitan area. It'd do even better at interior partitions in larger structures; before we close up a floor, load in 250 panels with mark numbers and precast-style erection tickets and then let the interior fit-out folks erect after that floor is weather-tight. Walls placed, utility connections made, structural links fixed, paint, trim, done.
I have no idea how to even begin analyzing this, but it seems very much untouched.
Has anyone tried to pencil out what the "ideal" cost-effective row home / attached single family setup might be, both in the dimensions of each unit, and perhaps the marginal cost of increasing block length? It seems like a potentially beneficial problem to solve, because they implicitly would seem to reduce redundancies in exterior walls and roofing, while (as far as I know) still allowing for fee simple ownership and the relative lack of financing complexities associated with that, if footprint lots are permitted by local zoning. And with land being one of the most significant portions of housing costs at the present (at least for 1-4ish unit scale residential, so it seems) attached row housing seems as efficient as it gets
The size of each unit in historic Mid-Atlantic rowhouses seems pretty clustered around 15-25 feet of frontage and 40-60 feet of depth, but I'm sure that could be an equillibrium that responded to financial incentives and construction practices that may no longer exist to produce the same result
In Israel it's common for contractors to build a large number of identical buildings next to each other. For example we live in one of about 50 identical single family homes, next to us are 15 identical 10 story buildings, and next to them 3 identical 25 story buildings.
I'm guessing this is a way to capture many of the economies of scale (1 design for all buildings, learning curves, buying in bulk), whilst avoiding any of the diseconomies.
I have believed for a while now that the biggest benefits to economies of scale would be from producing wall segments and floor segments in larger blocks. A standardized wall assembly with conduit embedded for electrical and plumbing routing. Then tip up the 20+ foot walls. the bigger the sections and the closer to final finish the more of that assembly can be done with assembly line or computer controlled tools. I think we hit a productivity limit with 4x8 sheets and loose 2x4s
This is very interesting thanks for writing it, but your analysis mixes up market and physics issues with policy issues - like permitting fees, the opinion of neighbors, whether you need stamped plans - which vary a lot more across towns than physics or market issues.
Thank you for your articles, very well done. Re this questions (why larger bldgs aren't much cheaper...), as you probably know, the rule of thumb (for most buildings) is that the bare structure (which is much of what you focus on) is only 20% of the cost of the finished building construction cost (ie, omitting land cost). The other 80% are interior finishes (finished floor, walls, fenestration...), MEP etc. These 80% scale almost linearly with floor area. On the other hand, I agree that the economies of scale not occurring is very puzzling - I'm still astounded to go by US sites and see stick-built construction still happening, even on 4 or 6 story apartments over concrete podiums. Its real piece-work. In northern Europe there are many more economies (eg, panelized construction - they tend to replace more labor with capital). Part of the reason is that, surprising to me, 2x4 wood framing is still the cheapest option in N. America for small-medium residential construction. Again, thanks for thoughtful pieces. (PS - would Munger live in that dorm? doubt it. the psych costs would appear terrible).
As Charles points out, most building costs are interior fit-out, and that's even more true for single-family homes where the structural system and design costs are a pittance. One of the greatest challenges for prefabrication methods is that they stall out at the structural level, maybe with some architectural finish. We don't do much prefabrication for either MEP or interior finish work. That's true of precast concrete and tilt-up systems, panelized wood structures, both cold-rolled and light-gauge steel assemblies, glazing modules...
What I'm not really sure of is whether we *can* do that economically; talking to the professionals, it seems no one is trying.
For example, is there an impediment (genuine or code-based) to creating a fully panelized, factory-assembled system for single-family home walls, where utilities "plug into" standardized connections in a wider building system? For single-family homes, such a system would seem to be within reach if I'm building several hundred similar family homes in 8 sub-divisions in a metropolitan area. It'd do even better at interior partitions in larger structures; before we close up a floor, load in 250 panels with mark numbers and precast-style erection tickets and then let the interior fit-out folks erect after that floor is weather-tight. Walls placed, utility connections made, structural links fixed, paint, trim, done.
I have no idea how to even begin analyzing this, but it seems very much untouched.
For residential buildings, upstairs neighbors are also a negative (i.e. diseconomy).
Has anyone tried to pencil out what the "ideal" cost-effective row home / attached single family setup might be, both in the dimensions of each unit, and perhaps the marginal cost of increasing block length? It seems like a potentially beneficial problem to solve, because they implicitly would seem to reduce redundancies in exterior walls and roofing, while (as far as I know) still allowing for fee simple ownership and the relative lack of financing complexities associated with that, if footprint lots are permitted by local zoning. And with land being one of the most significant portions of housing costs at the present (at least for 1-4ish unit scale residential, so it seems) attached row housing seems as efficient as it gets
The size of each unit in historic Mid-Atlantic rowhouses seems pretty clustered around 15-25 feet of frontage and 40-60 feet of depth, but I'm sure that could be an equillibrium that responded to financial incentives and construction practices that may no longer exist to produce the same result
Very interesting!
In Israel it's common for contractors to build a large number of identical buildings next to each other. For example we live in one of about 50 identical single family homes, next to us are 15 identical 10 story buildings, and next to them 3 identical 25 story buildings.
I'm guessing this is a way to capture many of the economies of scale (1 design for all buildings, learning curves, buying in bulk), whilst avoiding any of the diseconomies.
Your Works in Progress link is producing a message about an expired certificate. It could just be my browser though.
> since the costs of climate control would (theoretically drop).
I think this is a typo.
Also the "Why are Skyscrapers So Short?" link appears to be broken.
I have believed for a while now that the biggest benefits to economies of scale would be from producing wall segments and floor segments in larger blocks. A standardized wall assembly with conduit embedded for electrical and plumbing routing. Then tip up the 20+ foot walls. the bigger the sections and the closer to final finish the more of that assembly can be done with assembly line or computer controlled tools. I think we hit a productivity limit with 4x8 sheets and loose 2x4s
This is very interesting thanks for writing it, but your analysis mixes up market and physics issues with policy issues - like permitting fees, the opinion of neighbors, whether you need stamped plans - which vary a lot more across towns than physics or market issues.
These are the best topics! Every single one illuminates my knowledge about construction and real estate. Amazing!