Friday, January 27, 2012

Friday Read: Mixed-Income Housing, Prussian-Style

In the response to Wednesday's post, commenter Marc mentioned the example of the Mietskaserne (roughly, "tenement"), the mid-rise courtyard apartment blocks that were built in the tens of thousands throughout central Europe during the 19th century.  Europe's answer to the tenements of the Lower East Side, the designers of these apartment buildings attempted to accommodate density with a degree of open space and residential with commercial and industrial uses.  Notably, they even aspired to integrate the wealthy with the poor and middle-class, all within the same buildings or cluster or buildings.

The Berlin city planner James Hobrecht, who played the role of Pierre L'Enfant, Andrew Haswell Green and Frederick Law Olmsted all rolled into one, had already distinguished himself from his counterparts in Washington and New York by incorporating existing roads into his plan, rather than blotting them out under an orthogonal grid.  His approach toward residential development, also, represented a quite different approach from the consensus emerging in mid-19th century London and New York, as this study describes:
"In a 1868 publication Hobrecht reveals his position on housing. He describes a situation in English cities where wealthy inhabitants would live in their villas in West-end districts. They would be completely separated - spatially and socially - from workers. Hobrecht rejects the English model with its strict spatial separation of classes on the scale of districts. Instead he illustrates the Berlin model, in which Mietskasernen play a crucial role. He describes the Mietskaserne as a multi-storey building with the following structure
of dwelling units and respective rent prices:
  • Floor IV: 3 units à 100 Taler
  • Floor III: 2 units à 150 Taler
  • Floor II: 3 units à 200 Taler
  • Floor I: 1 unit à 500 Taler
  • Ground floor: 2 units à 200 Taler
  • Basement or rear building: units à 50 Taler"
The study criticizes this arrangement for potentially exploiting the poor for rent, in substandard accommodations, and under the watchful eyes of the agents of the building owners and the wealthy residents on the étage noble, yet it's not clear, based on the facts given, whether the fourth floor or basement apartments actually paid for themselves (although it's said that the poorest residents often failed to pay their rent altogether).  Were the wealthier residents effectively subsidizing those in the basement and fourth floor?  Hobrecht's discussion of how he envisioned the residents living together, moreover, sounds like it was excerpted directly from The Death and Life of Great American Cities.

Berlin Mietskaserne.
As for the quality of the accommodations, Hobrecht's plans do concede a measure of comfort for the sake of affordability, a tradeoff which appears to reflect a market-based and realistic approach to affordable housing that contrasts with the garden city idealism of English and American planners.  Even so, certain pseudo-utilitarian fetishes of the day do make an appearance: an obsession with light and air, for one, excessively wide streets, and a requirement that courtyards be large enough for a fireman's wagon to make a u-turn.  Use-based zoning was absent, however, at least initially, with the buildings serving as residences, small-business incubators and manufactories all in one.

The study contrasts the positive and negative views of these apartments as being simultaneously "palaces" and "prisons", but even this dualistic interpretation contrasts with the almost uniformly negative appraisal of the tenements of Manhattan (although the tenants' own views probably differed from those of the most vocal critics).  Marc, in his comment, speculates that the design and quality of these dwellings may have played a role in the divergent attitudes of 19th century continental Europeans and Americans toward high-density living.  At the very least, this shows one alternative approach to planning, and to an integration of uses and incomes, that still has relevance for today's planners.

Wednesday, January 25, 2012

City Blocks: The Spaces In Between

A reader recently emailed me with an interesting question:
 "While I have a good sense of the design elements of the narrow street and attached housing for the front, I don't see much of what the back looks like.  Is it better to only have windows on one side and have a separate residence with another shared wall, This seems the most effective but do people want sunlight from both sides? Are they using private backyards, an alley, or both together? ... Are there shared courtyards and should these have grass? Isn't that similar to 'green space'.
The story of what has taken place in the spaces between blocks is an unexplored chapter of urban history. Every now and then particular episodes have been covered, with writers generally critiquing the process by which open space in block interiors yielded to new buildings.

Lewis Mumford in The City in History described, with disapproval, the process by which rear yards in medieval cities were filled in with houses and apartments as urbanization intensified in the late middle ages. Haussman has been criticized for permitting landlords to construct "airless and crowded tenements" behind the elegant street facades of new Parisian apartment buildings.  Others have lamented Barcelona's failure to maintain the interior of blocks in the Eixample as green space.  Even the modest concept of the mid-block laneway house has become a subject of controversy.

The process of block subdivision, infill and densification isn't an aberration, however, but instead represents the basic engine of urbanization in all times and places, at least where the process is not stunted or halted entirely by regulation.  It is more noticeable when it occurs in the context of an idealized plan, such as Haussman's for Paris or Cerda's for Barcelona, but the same process is at work in thousands of other less noteworthy cities.

A 1615 drawing of Paris shows the process well: the blocks outside the old city walls, to the left, are thinly lined with rowhouses backing onto ample gardens, while inside the wall, blocks once similar have been carved into much smaller pieces by narrow streets until, at farthest right, interior garden space has almost vanished.

To return to the question that was posed, what is happening to the interior of the blocks during this process?  In the unplanned city, the process of subdivision generally proceeded until an optimal ratio was struck between streets and accessible block space. Driving the process is the principle of maximization of land value, where a new street was justifiable only to the extent it increased adjacent land values in excess of the space lost to right of way, bearing in mind that ground floor, street-front rents are almost invariably command rents much higher than those on any other floor.

The result, in almost all cases, was a proliferation of small, or at least narrow blocks, generally not much more than 150 feet wide, and frequently a lot narrower.  The image below of Bordeaux shows the end result of this process, with the block partially outlined in yellow, for scale, measuring 115' by 380':


With blocks of this width, few apartments are entirely buried within block interiors, lacking street-facing windows or exits, while light shafts admit sunshine to inward-facing rooms. There are no rear alleys and no yards, yet most apartments do receive natural light from at least two directions.  Larger blocks frequently incorporate interior courtyards.

The same pattern can be imitated by planners, as in the case of Taipei, with the outlined block, one of the largest in this neighborhood, measuring only 75' by 200', with a small space left between back-to-back apartments.  Streets are generally about 16' wide, with wider arterials at intervals. The amount of street frontage within a five-minute walk of this block must be phenomenal, providing a resident with access to many dozens of businesses steps away from his front door.


Due land value gradients, block size in the unplanned city generally shrinks toward the center.  With planned cities, the opposite pattern is often seen, due to a tendency for grid-makers to overscale city blocks at the outset. leaving too much space too far from the street grid (a decision compensated for, in some places, by the introduction of alleys).  Mexico City and Salt Lake City were laid out centuries apart, but in both cases a clear transition toward much smaller blocks is evident beyond the original gridiron as the economic picture came into clearer focus and the benefits of narrower blocks became more evident: in Salt Lake City (below right), from 720'  to 380' square blocks, and in Mexico City from blocks 270' wide to 120' wide:


The dimensions for the Mexico City blocks closely approach those which Nathan Lewis has proposed in his writings: see How To Make a Pile of Dough with the Traditional City 5: The New New Suburbanism (scroll to the end for diagrams).  Blocks of this width can accommodate almost any type of structure with little wasted space: single-family homes with modest yards in Tokyo, rowhouses with small rear courts in Philadelphia, large apartment buildings in Barcelona.  Even skyscrapers are entirely possible.  At 200' feet and above, however, the width of the narrowest blocks in most American grids, even the largest townhouses struggle to fill up the space.

Ultimately, the question of what is going on between blocks depends in large degree on the size and shape of the blocks themselves, and that is a topic on which the experience of cities from around the world have a lot to teach us.

Related post:
Blocks of New York

Other discussions:
Why are Blocks that Size?
The Variety of American Grids

Friday, January 20, 2012

Friday Read: New York's Early Height Laws

In the past I've written about the effects of the design of New York's grid plan and particularly its division of blocks into lots of 25' by 100', dimensions that were, in the mid-19th century, incompatible with the construction of well-lit or ventilated apartment buildings. A 2003 paper published by the Cato Journal echoes these observations, but focuses on a less well-remembered aspect of New York's planning history: the building height act of 1885.

Although this law was overshadowed by the Tenement House Act of 1901 and even more so by the Zoning Resolution of 1916, the 1885 act established a precedent on building heights that would persist until the 1961 amendments to the New York zoning code and beyond.  The law, simple enough, capped building heights at 70 feet on narrower streets and 80 feet on the wide avenues – limits comparable to those found in European cities of the time – yet, unlike the laws of those cities, the New York ordinance only restricted residential buildings, leaving commercial structures, and hotels, unregulated.

The effects of this selective restriction are described:

"First, economic rents accrued immediately to owners of existing tall residential buildings due to the restriction of potential competition from the new buildings that would not now be built. …

"Second, developers of apartment buildings found themselves unable to compete with commercial-building interests for Manhattan building lots offering appeal either as residential or commercial sites. Prior to 1885, large apartment buildings had been constructed that not only competed with contemporary office buildings in their ground coverage but exceeded them in height.   This era now ended. …

"Finally, height restrictions helped prop up the tenement system by inhibiting the development of the tall apartment building as an indirect— and over time a direct—competitor. By preventing the development of tall residential buildings in Manhattan, height restrictions derailed a natural market process that would have lowered rents and increased quality beginning with the middle-class portion of the tenement market."
Although the law did not impose Euclidean zoning on the city, in effect the law zoned all buildings over the height limit as commercial, since these could no longer be converted to residential use.  In doing so, it anticipated the explicit zoning by use in the 1916 act.

The law also had long term implications for the arrangement of jobs and residences.  The observation that apartment buildings taller and larger than office buildings were beginning to appear in the city makes sense when we consider that office buildings have many more workers, per square foot, than apartment buildings have residents.  To fill the new office towers by pedestrian traffic would have required an even greater number of nearby residential towers.  With these prohibited, it would fall to mass transit to link the increasingly tall office buildings of lower Manhattan and Midtown with the residents of the six and seven story residential buildings spreading northwards up the island.

The strict height limits of the 1885 law were moderated somewhat in 1897, when New York moved to permit residential buildings of up to 12 stories on all streets.  This arrangement quickly gave way, in 1901, to the familiar "1 and 1/2" height formula, which restricted apartment buildings to the width of the adjacent street plus 50 percent.  It was this limit which was incorporated into the 1916 zoning resolution for residential areas, and which persisted for the most part until 1961, severely depressing the market for apartment construction in the city – with the exception of the avenues, where the formula yielded buildings of profitable dimensions (the claim of New York's planning department that the 1916 code could have accommodated 55 million seems questionable to say the least).

Perhaps the greatest irony of all is that the city's ultimate remedy for the so-called tenement problem – the form of the tenements and their persistence a direct consequence of city planning decisions dating back to 1811 – was to demolish tenement areas and replace them with high-rise apartments of the sort that the city had barred the private sector from building in the critical years from 1885 to 1930.  As New York reexamines its zoning code, will the city pause to consider its 200 years of planning history candidly and critically to see what lessons might be learned?

Related posts:
Blocks of New York
Living Space, Working Space and Centralization
Height Limits: The Forgotten Debates

Friday, January 13, 2012

Friday Read: Affordable Housing, Filtered, not Subsidized

Despite the enormous amount of attention that has been given to the subject of affordable housing in the recent past, and the wide array of programs and initiatives in response, including low income housing tax credits, inclusionary zoning and voucher programs, to name a few, there's been relatively little formal study made of filtering, the process in which housing gradually declines in relative price, and in rental value, as it ages. To fill this gap, economist Stuart Rosenthal has released a new paper in which he concludes that filtering, more so than many place-based subsidized housing programs, can provide a viable long-term source of low cost housing.

Using data from the American Housing Survey, and examining the income of owners and renters at various points in time, Rosenthal constructed charts showing the change in inflation-adjusted incomes of residents for homes of ages ranging from one to 86 years:


Based on the observed rates of downward filtering, which range from .5 percent per year for owner-occupied housing to 2.5 percent per year for rental housing, Rosenthal concluded that "policy makers should take seriously the market’s ability to generate lower income housing, and especially in the rental sector of the
market."  The study noted that differential rates of filtering by city and state also have implications for the efficient apportionment of federal housing money.

An interesting finding was that homes over 50 years old showed an increase in owner or renter income, even though the survivorship effect was controlled for by following individual houses over time.  Rosenthal hypothesized that older homes may have intrinsic structural qualities that make them more desirable to  buyers.

That hypothesis tracks my own finding which I discussed here, suggesting that the process of filtering, at least for initially high-income older neighborhoods, was a temporary result of rapid suburban dispersal in the 1940-1970 time period, and since that time has been steadily reversed.  On the other hand, some persistently disfavored quarters have remained poor throughout the 1940-2000 time period, showing little evidence of further decline or imminent gentrification.

Is filtering, then, partially an artifact of a temporary and unsustainable process of suburbanization, with gentrification representing a reversion to a pattern determined more by centrality of location and quality of housing stock or urban environment rather than size or newness of the home?  What that trend holds for the future of market-driven affordable housing is not entirely clear, although dire prognostications of coming suburban slums have appeared repeatedly in recent years.

One could hope that increased demand for existing urban neighborhoods might at least translate into a  process of steady upgrading and intensification of urban areas, yet the opposite seems to have occurred in most places, with residents holding fast to exclusionary zoning measures designed to impede new construction or conversion to higher value uses, even to the extent of zoning out larger single family homes.  The ultimate antidote to urbanism and densification, zoning, remains strong even nearing its 100th birthday.