It is common in construction to normalize per m2 floor: such as e.g. energy, material, ventilation needs, or even costs and investments. But when it comes to the national requirements for housing in fact you should not calculate per m2 floor. Because it’s not about the impact per 1 m2 , but about the impact of the whole construction setcor and even the absolute impact of all the m2 floors in the Netherlands together. After all, maybe we should reduce constructed m2’s, since we might be able to produce and use them effectively, but there is just no ‘budget’ for it, in terms of energy , materials and CO2. [see previous contribution 
Nevertheless, for the construction sector , to compare solutions, it is of course interesting to see what effect 1 m2 of floor actually has. But moreover, in that case it must first be clear what is actually meant by it, “per m2 “?
That is, as a reference, such as kWh/m2, kg/m2 euro/m2, and/or CO2 emissions per m2. What are we actually measuring? And especially: over which m2 floor? There are many definitions of that, and they are used differently in evaluations. Therefore, some thoughts on calculating per m2, starting with which m2, of a house. The essence is to create a room, a ‘liveable inner space’, to live in. So to make a number of m2 available, which can be used , lets say ‘rented’, and which need to be heated in winter.
But a room needs a door to enter. So you can reason, every time that door opens you lose heat so a draft portal is needed. Suppose that is 1 m2 in size. So where does that m2 belong to? It’s pretty useless to the user, is only for energy saving reasons. However to reduce operational energy loss another energy burden has been introduced: an extra amount of embodied energy, for door and portal. So, the impact of that extra m2 and its embodied energy needs to be taken into account . But not averaged including that extra m2 , which in terms of functionality, is useless: it is not ‘rentable’, so to speak. It is therefor the impact of the whole, the construction of the space, heating the space, and building a draught hall, and all that averaged per m2 of original -usable- space, so without draught hall. After all, it’s about providing functional, “livable”, rentable living space.
If you then put two or more rooms cq ‘apartments’ next to each other, you create some advantage, by joint use of a partition wall: that limits the material use, and heat loss on that side. The comparison becomes more favorable. Moreover, that saves some land area, because the portal can be shared and both have still full access to, for example, solar energy on their roofs. That is not hindered. The impact of a usable m2 becomes lower. Perhaps at the expense of some noise hindrance, but we will leave that aside for now.
What happens if we stack the two apartments instead of putting them next to each other? In terms of energy losses, this does not make much difference for well insulated rooms: there is the advantage of shared roof and a shared floor instead of shared wall, and if constructed in wood , not too heavy, the foundation will also be a bit cheaper overall. But with stacking you need stairs and possibly galleries. So these add up to the total invested material impact, just like now a double draught portal is needed. While the useful m2’s do not increase. (More about land use later).
The distribution of the total impact over the useful functional m2 does lead to a higher impact per useful m2 for the stacked form compared to the terraced form. It is clear that care must be taken with what is and what is not included in calculations per m2, in all kinds of tools.
Of course, most homes do not consist of 1 room. At least not in the rich countries. They have several rooms for separate functions. But that also means more traffic space inside, which is basically useless. Useless, meaning not rentable: It becomes clear if you imagine a student house: The rooms are useful and rentable, the rest is not. Even a kitchen in such a house can only be seen as a bonus.
A nice example of how this works out in a house are the Agrodome houses in Wageningen. At the time, they were built very sustainably, and mainly with bio-based materials. But the impact of the materials and energy divided over useful (rentable) m2 is still high. One reason for this is that the houses are very narrow and deep, requiring a relatively large amount of traffic space. If you then divide the total impact per net useful m2 (the rooms themselves), it still ends up high (there is of course some over-space of course, which might be handy for parties for example….). (see ill. on home page message)
In the same way, you should calculate for apartment buildings: and then corridors, staircases, galleries, basements, sheds all count and add tot the total environmental impact, but do not qualify as as m2 in the normalization, the environmental impact is only allocated to the net usable m2. Of course this rarely happens in practice, in both houses and apartments almost always all m2 are usually counted as useful.
Not to mention parking garages, under apartment buildings. Completely useless, in terms of housing functionality. Besides, a garage is a relic from the 20th century, and can be discarded as an archetype. (cars don’t rust anymore) A parking garage under a building has a very high impact by definition, because it is underground and therefore contains of an enormous amount of concrete and steel. The impact of this must be taken into account as a whole, and divided over the useful habitable or rentable square meters. But better not to build parking garages.
That also applies to the latest hype: extra (heavy) balconies with trees…. I wrote about that earlier.
In all of this, energy consumption, and its standardization, plays a special role. Because that is also usually averaged over all m2 , not just the ‘useful’ ones. For example, in the Netherlands the so-called BENG regulations for buildings , requires a maximum number of kWh per m2. But which m2 then? Is that only the living area, all rooms, or eg also corridors, stairs, laundry rooms etc? Officially, living area is defined as ‘user area’: where almost everything is included, except spaces lower than 1.5 meters (under a sloping roof eg), and voids or stairwells if larger than 4 m2. But that hardly ever is the case.
So almost all m2 of a house count , but then how many m2 , how big is a house? That again is not defined….
By setting a standard per m2 over ‘all’ m2, you are implicitly demanding that the whole house must be optimized with insulation and equipment, or renovated, to get below that energy standard. With all the impacts of additional material and embodied energy that entails. But what if someone adapts behaviour and limits himself to heating the living kitchen, and not the rest of the house? Then he could, in principle, already meet the requirement without insulating anything. Because averaged over the whole house, that still works out well within regulations. (Only the required calculation does not allow this).
It can also be, that a core of the house is optimized and insulated, say that kitchen, and as such meets the actual requirement per m2, leaving the other spaces not insulated. Of course, in terms of impact, that is a much better solution than insulating the whole house as if it were freezing all year round, saving a lot of embodied energy ( and operational energy as well). But then this does not meet the regulations, since theoretically if you would want to keep the whole house warm, which is largely uninsulated , then the energy demand would be far above the standard.
Anyway, a calculation in kWh per m2 for operational energy of course totally fails to take into account the material impact per m2 , which can also be expressed in kWh (the manufacturing energy or Embodied energy). Those two, energy and material, you can not in fact not treat separately. For this reason I have previously argued for a additional ‘BENG-requirement, a Maximum of embodied energy (in kWh). It must be said, in the Netherlands we have a legal requirement called MPG ( material performance of a building). However the embodied energy emissions of materials are only 1 of many parameters in the MPG, packed and weighed, and as well averaged per m2….
While a house with more m2’s always has less impact per m2 than comparable smaller houses… so the averaging encourages building bigger.  (Incidentally, the MPG is also averaged over the lifetime. But that is calculating yourself rich, as far as materials are concerned the impact is direct, at delivery.) .
Still not included is material depletion and regeneration.  For that you can also add a kg/m2 requirement …. 
So calculating per m2 of floor of a house can still inflict a huge waste of energy and material (to bring whole house within the standard). Yes, it’s called comfort, I know, but now is not the time to improve comfort but to reduce environmental impact.
Calculating per m2 can be useful to compare the same kinds of spaces, but not as a measure of environmental impact, not as a measure of absolute reduction, and so certainly not in terms of regulation. That linkage needs to change. The current one (in NL) leaves room for the wrong solutions: that may create ‘efficient m2’s’, but with that, not necessarily absolute environmental impact reduction.
It was understandable at a time when the limits to energy and especially materials were not yet so dominant, and one wanted to guarantee a minimum quality of a whole building (in this case with regard to energy demand in the winter period). But now it is about the absolute impact in energy, material and CO2 load. Even other aspects come into play, such as that a standard per m2 also hides the discussion about a maximum of living m2 that we occupy per person [see 7]. Currently that’s not an issue, someone who builds a house can build as big a she wants, as long as its within a standard per m2. And bigger is profitable per m2…. That doesn’t really help. So it has to be done differently. For example, an absolute budget per house, large or small, efficiently designed or not, regardless if fully or partially heated. Like for example calculating with an absolute CO2 budget, as argued in the previous contribution.
PS1 In addition, the (material) impact calculation of a m2 floor space should actually also include the related secondary functionality: roads and other infrastructure in m2 per m2 house, which actually only exists because there is a house constructed, and whose impact should therefore be added. A house does not stand alone! But about that another time. ( or see )
PS2 In general, caution is therefore required when using calculations to m2 , and interpreting them*. By the way, to get some insight into this matter, I practice with students the following requirement: The maximum energy demand (also of material investment!), (or so the ‘limited budget’), is linked to generation: the operational demand plus the investment (in EE) may not exceed the own building-related, renewable generation within a certain period, or number of years.
Then you link the m2 performance and simultaneously the number of possible m2 to be constructed on a given building plot (both heated and unheated m2’s) to the on-site capacity to be delivered on or to the building, to provide the energy for this. For both operational and embodied. That make sit m2 independent, that is, determines the m2 potentially to be constructed. And that requires a thorough consideration to create a balance between energy demand, energy generation, and energy investment (in the materials for both ‘reduction’ and generation). See among others here  .
 co2 budget per building: http://www.ronaldrovers.com/a-co2-budget-limit-for-housing/
 trees & balconies: http://www.ronaldrovers.com/vertical-frarming/
 copper 8 sybren https://www.copper8.com/aanscherping-mpg-stuur-sterker-op-maatschappelijk-belang/ (Dutch)
 embodied emissions are direct: http://www.ronaldrovers.com/embodied-energy-emissions-is-now-not-in-2050/
 maxergy 3.0 : Soon to follow. ( in Dutch: http://ronaldrovers.nl/de-hernieuwingstijd-van-alle-bronnen-maxergy-3-0/ and a brochure: https://bewustbodemgebruik.nl/wp-content/uploads/2021/08/Volhoudbaar-landgebruik-BROCHURE-2021-1.pdf )
 m2 living area per capita: http://www.ronaldrovers.com/m2-living-per-capita-in-europe-and-consequences/
 roads and sidewalks : http://ronaldrovers.nl/exploring-sidewalks-impact-similar-to-housing/
 climateneutral per m2, and m2 pv/m2 floor