[N]ew entities called "efficiency utilities" ... would pay for efficiency upgrades in order to bring an existing building in compliance with the limits. Owners/tenants would pay for these improvements via a monthly bill and, though they would be part of the building, the improvements' cost wouldn't require "recouping" by the owner in the form of rent hikes or higher a sales price. A particular unit would simply have a particular monthly cost for "efficiency" like it has a monthly cost for heating.Of course an efficiency utility wouldn't just cover insulation, caulk and new windows -- it would cover heating systems, appliances, shower heads, etc. A further advantage to a utility model over the financing model that Leonhardt discusses -- the idea of adding weatherization costs to homeowner's property tax bills -- is that it addresses the fact that weatherization doesn't lend itself to one-size-fits-all solutions. As Leonhardt observes, the complexity of retrofitting old homes is enormous:
And like electric service, the "efficiency" bill can be stopped - if an apartment sits unrented, for example. Because both the utility as well as the bill itself could be subsidized in various ways it would, according to Lipow, remove a major stumbling block to making improvements in existing buildings. For the record, an efficiency utility could cover the costs associated with:
What share, say, of Midwestern homes built before 1950 could use more attic insulation? How quickly would the insulation pay for itself on average? Every home is different, obviously. But without any reference point, many people won’t be confident enough to plunge into a project.Even if they don't ultimately perform the work themselves, a utility would have the scale to provide the expertise as well as the data for what particular homeowners should do. Obviously, this kind of program would go beyond what any stimulus bill is likely to enact. But if we want to make efficiency a goal unto itself, a utility model -- not to mention per square foot emissions limitations -- is the way to go.