Chairman BOLLING. Our next witness will be Mr. Richard H. Rosen, President of Urban Systems, Inc., of Boston, Mass.

Mr. Rosen, we are glad to have you with us and grateful that you came, and you may proceed as you wish.

STATEMENT OF RICHARD H. ROSEN, PRESIDENT, URBAN SYSTEMS, INC., BOSTON, MASS.

Mr. ROSEN. Thank you very much, Mr. Chairman.

I think before beginning my opening remarks, I would like to refer in part to some of the comments that Congressman Reuss made earlier in his questions to the other gentlemen.

It is my experience that there are a number of confusions which mitigate against sensible analysis of the entire urban problem, especially with respect to housing. The first of these is a complete misunderstanding of the whole concept of economies of scale. In this respect I would like to refer to the point which Congressman Reuss made and to attempt to attribute the areas of potential economy to their actual source and to refer perhaps even more specifically to the question that he raised with regard to single-family homes.

First, it seems to me that economies of scale can be derived from three sources. The first source is just a pure economy owing to quantity purchasing power. If you go the General Electric Co. small appliance or housing appliance division and say I will buy that production run for a given period of time, you will probably enjoy a substantial cost savings. Everyone likes a reduction in uncertainty and everyone likes to operate at an optimum level of production. The general efficiency of the large single family homebuilders in the United States whose costs range from my experience between 6 and 11 dollars a square foot for pretty damn good housing is significantly due to their ability to enjoy mass purchasing economies.

I did not have an opportunity to read the testimony that the gentleman that you had here from Levitt gave, but I have enough personal experience with Levitt and other individual homebuilders to recognize that at least half of any possible economies derive from that effect specifically.

Secondly, I think that the single-family homebuilding industry, the large builders (I am not talking about the 106,000 small homebuilders in the United States, most of whom build fewer than 15 or 20 houses), like Levitt, like Ryan Homes, like some of the others who enjoy substantial managerial efficiency largely as a function of being able to afford professional management in terms of job organization. It is clear that professional management is really the second source of saving independent of any particular technology.

Now, to the extent that a particular technology might enhance that saving, it can be of real interest. I think from some of the remarks that appear in your own documents of this committee of the April proceeding and certainly in what has been published in the literature in general that the opportunity for cost saving in systems per se as opposed to job organization and material supply are a function of the size of the job itself.

What I am going to say here in the next few minutes is really an indictment of potential large-scale building systems development in

ing rules, so that where land is cheap, generally speaking, density is United States. It will be a description of natural inefficiencies owing to very primitive industrial organization of the housing industry in this country.

I also would like to preface my opening remarks with a feeling that there is not going to be any real housing industry in the United States unless we do something about stabilizing that industry from interest rate fluctuation. I think that the cyclical variability which affects construction in general and housing in particular causes such a high level of instability in the industry that it makes it almost impossible for any kind of managerial overhead to be developed sensibly. Representative BOLLING. May I interrupt you at that point to get your suggestions as to what kind of things should be done, what type of approach should be used to insulate the housing market from the cyclical impact of interest?

Mr. ROSEN. OK. Well, one opportunity might be for the Government to produce essentially a housing bank which would agree to purchase a certain number of housing units each year above and beyond some given average number, and I am not specifically sure now that formula could be determined off the top of my head. But it seems to me that there is an opportunity in that respect to insulate the industry. I think a second way is to provide some capital security. One of the problems that Mr. Price referred to which I also refer to in the testimony that I shall deliver here, is the problem of land acquisition. It is clear to me that you are not going to get any extensive largescale development in the United States unless in one way or another we make it easier for developers to assemble parcels of land large enough to develop economies on site. There is a natural reticence on the part of any developer, who tends to be capital poor under the present structure of the housing industry, to make these kinds of investments. Essentially the problem of the job of land assembly is largely one of option trading as a means of essentially minimizing capital requirements. What this shortage of land does is to present so much instability to developers that they have no opportunity to make manufacturing investments per se which might in effect lead to the adoption of a particular industrialized system.

Representative BOLLING. Again, do you have an idea as to what kind of approach?

Mr. ROSEN. There is a lot of talk, about air rights for example, but it seems to me that a large number of potential sites in the future in urban areas are going to be air-right sites, and there is no real mechanism in many places to use these sites. Where these mechanisms. exist they are in the hands of municipal authorities or quasi-public bodies or perhaps under the jurisdiction of the bureau of public roads or the State departments of public works or what not. It seems to me that all the air rights ought to be in the public domain because they are fundamentally a public good, and there should be some rational way of distributing these air rights to the public. I think that is one source of potential sites and I think that that is a real source which could be relatively insulated against any speculative uncertainty that is generally associated with land prices.

When I conceived of viewing housing as a production process, really finds that the cost of land in any one area is directly a function of zon

ing rules, so that where land is cheap, generally speaking, density is low. Where land is high, density is high. I did a study a year ago for the New York State Urban Development Corp. which looked specifically at opportunities for site development in Metropolitan New York and I found that the economic theory which says that prices should be equated at the margin for land of given use was in fact true if you just translated the density differences existing in one part of Metropolitan New York to another. So that there really was no effective difference in land zoned for one purpose in Metropolitan Manhattan, the Bronx, all the way out, 50 or 75 miles on Long Island and up into Westchester County. I think this is very interesting because it effectively confirms that the change in land prices is a function of the change in rules associated with the use of the land itself.

Now, with respect to one other thought I have as far as specific recommendations are concerned, it is my experience, again as a result of dealing with a large number of our clients, who represent several of the major corporations of this country, the prime inability, or the prime lack of willingness on the part of companies or firms to adopt new technology is that there is a high level of uncertainty associated with it, and managers of large-scale institutions are risk adverse. To the extent that they are risk adverse that means that any potential saving which might accrue as a result of the use of an industrialized technology must be substantial (it has to represent a substantial saving, maybe 30, 40, or 50 percent) before any manager in a large-scale institution is going to risk his job for the sponsorship of that system.

Now it is clear from the European experience that these savings will not be much more than 20 to 30 percent, perhaps 35 percent, once well developed. Well, once developed in the case of the European system has meant for example in the Czech situation 15 or 20 years and in some of the other situations 10 or 15 years. So it is unlikely that you are going to get any natural adoption by companies of industrialized systems, given the relationship between risk and uncertainty and the relative level of available cost savings with respect to the time period required for its realization.

I can allude specifically to a very large client of ours who is faced with this dilemma right now. They will, I am sure, enter the housing industry not with a true system per se so much as they will enter the industry with their managerial capability and financial resources, because with 8.5 percent, 9.5 percent interest, the ability to offer all kinds of innovative financing gives one a substantial comparative advantage over someone who certainly has no access to that kind of financing.

Well, with those few statements in mind, I would like to introduce myself, perhaps a little bit more formally, and say a couple of things with respect to what I feel is a sensible way of viewing the housing problem in terms of formulating useful public policy.

I have the good fortune to work along with John Collins, who was mayor of Boston and had some success in rehabilitating the city of Boston when he was employed in that responsibility. As you know, he is now associated with MIT and is chairman of the board and very active in the management of our company. In addition to my duties as president of Urban Systems, I also teach at Harvard University.

When I conceived of viewing housing as a production process, really as a means of developing sensible policies to encourage the adoption

of technology in housing, I noted that there were two ways to generate improvement and efficiency. The first was to change the relationship of inputs capital, labor, and materials-while the second was to determine the constraints which operate on the production process and develop ways and means of relaxing these constraints so as to effect an increase in output and a decrease in cost.

I discarded the first, a purely technological view of the problem, as being inadequate because housing is much less a problem of technology than it is a social process. Many people ask why we can employ technological resources and systems analysis to get us to the moon in a relatively short period of time and cannot similarly apply these resources to the problem of efficiently housing the people of the United States.

The constraints on the moon flight were almost exclusively resource restraints: money and trained personnel. Economists have known for a long time that an increase of resources in a particular area will relatively quickly develop the required technologically competent people to earn the high wages.

Improvements in the housing process do not depend on the commitment of resources alone. They depend on the ability of the Government to provide sufficient incentives along with the required resources to encourage the people who comprise the political and social institutions, whose behavior constrains efficient housing, to change their actions to those which will enhance efficient housing development. And I think we have some models for these to which we can refer. I think that given our specific objective here of really reorganizing the industry, allowing the industry to operate within a market context and efficiency basis, that we really have to look almost completely to a system of public policy which examines the incentives that are opened to the various parts of that particular industry.

I feel that analysis of housing must proceed from the constraints that impede the production of low-cost housing and inhibit the development and application of technology to improve efficiency. At this point I might say that if there were so many opportunities for tremendous cost saving, at least a few of the 106,000 firms who make up the homebuilding industry in America would be smart enough to realize that there were these improvements in efficiency possible. Given a number of reasonably intelligent and perceptive people believing that these improvements do exist and finding evidence that there are opportunities for cost saving, one must look almost exclusively to the constraints which impede their adoption and hope to develop the incentives to relax those constraints.

Now the constraints to which I will refer not only impede production and inhibit technological progress, they also act for the substantial variability of the cost of various aspects of the housing process because of their local character. In the course of this testimony I will refer to the impact of the very local nature of housing in many respects and what it does to one if one is really concerned with industrializing that process and in what the trade-offs are both from a policy and an institutional basis.

The work that we have done over the past several years has identified nine constraints as crucial. The first is industrial organization of the housing industry which I have mentioned in some of my earlier

remarks. The second, and a very important constraint, is the inability to interpret sensibly the preferences and behavior of potential homebuyers and users. We do not deal very well with obvious preferences of people. In Kansas City, for example, there is a housing project which is a model for the benefits of homeownership, one that was established by the Reynolds Co. It sits adjacent to one of the most difficult public housing projects in America. And yet at the same time the people who live in the Reynolds project are model citizens, although they are for all intents and purposes not very different from those who dwell in the public housing project. I argue that because of the homeowner opportunity, this particular project, the Reynolds project, provides a substantial change in the behavior of persons.

It is clear on the one hand that we do not bother to deal with the fundamental preference, homeownership, in any kind of useful way, but beyond that, we do not deal with any of the other preferences. We do not really know whether people want a lot of variability in the architectural exterior or whether they want a low rent as a function of low variability. We do not know, for example, whether or not people who live in houses want cheap furniture as part of the housing opportunity as opposed to paying substantially high costs for nonmassed-produced furniture and have substantially higher interest rates on that kind of furniture purchases, for example, than presently accrue to housing even though there are high-interest rates for mortgages.

The third constraint, which I have mentioned before, is the problem of the multiplicity of ownership of available sites and the problem of site assembly, and I think you are all sufficiently familiar with that so I will dispense with any additional remarks in that respect.

The fourth is the inability of the design profession to deal with building technology on a performance basis and similarly provides a substantial constraint. I have recently had an opportunity to work once again with a dozen of the leading architectural firms in the United States on a major project. It is completely clear to me that the present training of architects precludes their being able to work in any kind of a useful fashion with a systems approach to building. They have no sympathy with it. They have no rewards and they have no incentives to participate in this particular respect.

I also want to remark in the next few minutes about the fifth constraint, inefficiencies of the housing marketplace itself. I think the fact that there are diverse preferences, and that there are a large number of housing consumers, and that their needs are not uniform, make it very difficult to use a truly rigid system in any American urban area. Some research which I have been conducting with some of my colleagues at Harvard this past year revealed explicitly the dimensions of the problem of industrialized housing in typical metropolitan areas. Mr. Price's comments that there are only really two markets in the United States that lend themselves naturally to industrialized housing I think is a very accurate one, because in Boston, for example, where the number of housing starts per year is 10,000 to 15,000 a sensible size factory to provide an optimum level of production would need to produce roughly in the vicinity of 2,000 to 4,000 units on the basis of the European experience. It is highly unlikely that one out of five housing buyers is willing to live in the same type of structural entity. So then we are talking about a suboptimization in terms of the