When we say that housing is not "affordable" in the nation's most desirable regions, what is actually at stake?
Housing costs in these areas not only reflect constrained supply side realities, they are also representative of changes in relative income levels and global ownership patterns. There's a global marketplace which closed access cities participate in to some degree - for both services and housing. As for income, the reigning marketplace efficiencies of general equilibrium, increasingly mean employment via core or peripheral capacities. And much of the high skill knowledge based core, clusters in what have become the most desirable areas. Meanwhile, problems with access to areas of highly productive agglomeration, have been exacerbated by demand based responses which often don't take missing supply into account.
Even though it would be a long term approach, we could alleviate some of these additional access costs, by allowing more use of high skill knowledge in other regions as decentralized, primary market organizational patterns. In other words, we would be creating more productive agglomeration, in part so that aggregate housing costs could be spread across broader marketplace patterns. The excessive centralization of today's most important knowledge use, is an underlying factor in the housing costs of closed access cities.
If non tradable sector time based product is structured solely as secondary market dependence (as is currently the case), the process leads to too much societal pressure, for access to productive agglomeration. Among the many examples in this regard, are healthcare practitioners who (understandably) choose to locate where not only is it most pleasant to live, but where income potential is greatest. Hence these groups have considerable incentive to locate in closed access areas where their income can also be maximized, via locally existing revenue flows. Yet it's a process which can lead to an incomplete general equilibrium, due to lack of production and consumption of time based services in a full regional economic context.
Another way to think about the costs of housing in today's closed access cities, is the Baumol effect, which importantly serves as a coordination mechanism for a subset of potential time based services (at multiple income levels) alongside more direct forms of wealth generation. I emphasized subset here, because local zoning and regulation, also limits the extent by which middle and lower skill levels can partake in these local Baumol effects.
One advantage of knowledge use systems would be that - like the Baumol effect - time arbitrage also achieves income smoothing for mutual local coordination, since time arbitrage allows local group time value to function in relation to itself. One difference between a defined equilibrium and general equilibrium, however, is that time arbitrage would acknowledge the vital relationship between local income/resource capacity and local non tradable sector costs.
A better understanding of this relationship, would make it possible for small groups to take full responsibility for local infrastructure requirements and asset formation. By coordinating economic time value in relation to itself, more groups would be able to take part in our most important knowledge based challenges, without the near necessity of locating in areas which are currently exposed to high housing costs.