95th Congress JOINT COMMITTEE PRINT 2d Session
MULTIFAMILY HOUSI 11975-200
PREPARED FOR THE U E
SUBCOMMITTEE ON PRIO
ECONOMY IN GOVERNMENT OF THE
JOINT ECONOMIC COMMITTEE
CONGRESS OF THE UNITED STATES
NOVEMBER 14, 1978
Printed for the use of the Joint Economic Committee
U.S. GOVERNMENT PRINTING OFFICE 33-674 WASHINGTON : 1978
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402
JOINT ECONOMIC COMMITTEE
(Created pursuant to sec. 5(a) of Public Law 304, 79th Cong.) RICHARD BOLLING, Missouri, Chairman LLOYD BENTSEN, Texas, Vice Chairman HOUSE OF REPRESENTATIVES SENATE
HENRY S. REUSS, Wisconsin JOHN SPARKMAN, Alabama
WILLIAM S. MOORHEAD, Pennsylvania WILLIAM PROXMIRE, Wisconsin
LEE H. HAMILTON, Indiana ABRAHAM RIBICOFF, Connecticut
GILLIS W. LONG, Louisiana EDWARD M. KENNEDY, Massachusetts
PARREN J. MITCHELL, Maryland GEORGE McGOVERN, South Dakota
CLARENCE J. BROWN, Ohio JACOB K. JAVITS, New York
GARRY BROWN, Michigan WILLIAM V. ROTH, JR., Delaware
MARGARET M. HECKLER, Massachusetts JAMES A. McCLURE, Idaho
JOHN H. ROUSSELOT, California ORRIN G. HATCH, Utah
JOHN R. STARK, Executive Director
Louis C. KRAUTHOFF II, Assistant Director RICHARD F. KAUFMAN, Assistant Director-General Counsel
LLOYD C. ATKINSON KENT H. HUGHES PAUL B. MANCHESTER
WILLuAM R. BUECHNER L. DOUGLAS LEE DEBORAH NORELLI MATE
THOMAs F. DERNBERG PHILIP MCMARTIN M. CATHERINE MILLER
WILLIAM D. MORGAN GEORGE R. TYLER
ROBERT H. ATEN CHARLES H. BRADFORD, Count STEPHEN J. ENTIN
MARK R. POLICINSKI
SUBCOMMITTEE ON PRIORITIES AND E ONOMY IN GOVERNMENT WILLIAM PROXMIRE, Wisconsin, Chairmas
SENATE HOUSE OF REPRESENTATIVES
LLOYD BENTSEN, Texas PARREN J. MITCHELL, Maryland
EDWARD M. KENNEDY, Massachusetts GARRY BROWN, Michigan
ORRIN G. HATCH, Utah JOHN H. ROUSSELOT, California
LETTERS OF TRANSMITTAL
NOVEMBER 6, 1978.
To the Members of the Joint Economic* Committee:
Transmitted herewith for the use of the members of the Joint Economic Committee and other Members of Congress is a study entitled "Multifamily Housing Demand: 1975-2000."
The views expressed in this study are those of its authors and should not be interpreted as representing the views or recommendation's of the Joint Economic Committee or any of its members.
Chairman, Joint Economic Committee.
OCTOBER 31, 1978.
Chairmnan, Joint Economic Committee, U.S. Congress,
DEAR MR. CHAIRMAN: I am pleased to transmit herewith a study prepared for the Subcommittee on Priorities and Economy in Government entitled -"Multifamily Housing Demand: 1975-2000." The study was conducted by Profs. George Sternlieb and Robert W. Burchell of the Urban Policy Research Center at Rutgers University.
In addition to analyzing the Nation's multifamily housing demand through the year 2000, the study discusses potential problems to meeting the anticipated demand. I am hopeful that this study will prove useful to Congress in formulating future housing policy.
The views expressed in this study do not necessarily reflect the views of the subcommittee members.
Chairman, Subcommittee on Priorities and Economy in Government.
OCTOBER 25, 1978.
Hon. WILLIAM PROXMIRE,
Chairman, Subcommittee on Priorities and Economy in Government, Joint
Economic Committee, U.S. Congress, Washington, D.C.
DEAR SENATOR PROXMIRE: Transmitted herewith is a study entitled "Multifamily Housing Demand: 1975-2000," prepared by Profs. George Sternlieb and Robert W. Burchell with assistance provided by William G. Rainwater.
This study was conducted under direction of and reviewed by Deborah Norelli Matz of the committee staff. The committee is appreciative to Morton J. Schussheim of the Congressional Research Service, Raymond J. Struyk of the Department of Housing and Urban Development, and John Pitkin of the Center for Urban Studies of the Massachusetts Institute of Technology and Harvard University for their comments and suggestions in the preparation of this study.
Joint R. STARi
Executive Director, Joint Economic Committee.
MULTIFAMILY HOUSING DEMAND: 1975-2000
Precis ------------------------------------------------------------ 1
Sequence of i e paper ----------------------------------------- 3
I. Who lives in rental housing: A market profile --------------------- 5
Housing and population characteristics: Household size by
structure ----------------------------------------------RegionaJ variation ------------------------------------- 6
Household type --------------------------------------- 6
Education -------------------------------------------- 8
II. Household formation and the future demand for multifamily housing- 9 Introduction ---------------------------------------------- 9
Household type and structure 9
The first approximation of future renter housing need ---------- 12
The attrition 12
The share of the rental housing market to be secured by multifamily 15
III. Supply and demand factors in multifamily 19
The physical demands of 19
Barriers to meeting 19
High rate of government subsidy and inflation-------- 19 Land zoning and high interest 20
Operating costs escalation and the fear of rent control-- 21 Cooperative and condominium conversion ------------ 23
1. The competitive financial disadvantage of multifamily housing ------ 26 11. The future of section 28
III. Future additions or reductions of the rental housing stock through
condominium conversion or subdivision of one-family units-------- 29 IV. 30
Digitized by the Internet Archive
MULTIFAMILY HOUSING DEMAND: 1975-2000
By George Sternlieb and Robert W. Burchell*
Using Annual Housing Survey data on multifamily structure occupancy by household type for 1975, together with number of household projections from the Department of Agriculture, a gross level of multifamily housing demand is projected. Allowance is made for a 2 percent replacement rate of the total stock as well as a 5 percent vacancy figure for new household demand. Assuming that: (1) there is no massive shift away from one-family ownership to multifamily units, and (2) the present level of conversions from one-family units to multiple occupancy continues to offset the trend of conversion from rental multifamily units to condominium status, then total future demand for multifamily rental units is well within current construction levels.
We are presently building one and one-half times the demand
of 416,000 units per year projected for 1975-80.
The demand from 1980 to 1990 decreases to 367,000 new units
This 'Is further reduced to 335,000 units annually for the period
The major determinants of future demand will focus on the scrappage rate of extant facilities and the regional shif ts of population.
The future supply of multifamily structures depends on an abatement of construction costs and interest rates, and/or massive levels of Government subsidy.
There is evidence that operating cost rises are challenging the financial integrity of multifamily structures. This is manifested by the HUD and private market mortgage delinquency and foreclosure rate&.
Policy focus for the future should emphasize the minimization of both construction and operating, costs rather than augmented delivery rates.
*Director and Research Professor, respectively, Center for Urban Policy Research, Rutgers University, New BrunswickX.J. Mr. Sternlieb and Mr. Burchell were assisted by William G. Rainwater.
Multifamily housing and indeed rental housing generally is the stepchild of Government attention to shelter requirements. While just prior to World War 11 less than half of all American households owned their own residences, the --positive relationship between housing acquisition costs, operating elements and incomes has permitted a vast expansion of ownership.
The massive shift from a predominantly renter society to one of ownership is. shown in exhibit 1. From 1890 to 1930 the proportion of total housing units occupied by renters stayed between the 52 and 54 percent level. The enormous toll taken in America's household incomes in the Depression is reflected by an abrupt increase to 56 percent renters in 1940. Forced saving of the World War Il years, combined with vigorously productive Government policies, made possible rapid shift to ownership after this period. In 1950, there were 4 million more owner occupied housing units than renter equivalents; by 1975, nearly two-thirds of all American households lived in, their own facilities. In the 5 years from 1970 to 1975, the United States added 7 million net new oxAmer-occupied housing units compared with barely 2. million units for renters-and this despite the very brisk upsurge in the latter as a function of Government subsidy implementation. In turn, rental housing, particularly large scale multifamily facilities, have been seen as a largely transitional provision for a good many of its occupants until they can reach the new appropriate standards-a house of their own.' Indeed, one of the major objections to the large central city was that this latter type of provision simply could not be met there-and that often occupants in such areas had no choice but the "inferior' I rental
OCCUPIED HOUSING UNITS BY TENURE (U.S. TOTAL, 1975) JUnits in thousands
Total Owner occupied Renter occupied
Year units Number Percent Number Percent1890 -------------------------------- 12,690 6,066 47.8 6,524 52 ', 2
1900 -------------------------------- 15,964 7,455 46.7 8,509 53.3
1010 -------------------------------- 20,256 9,301 45.9 0,954 54.1
1920 --------------------------- 24,352 11,114 45.6 3,238 54.4
1930 -------------------------------- 29,905 14,280 47.8 15,624 '52.2
1940 -------------------------------- 34,855 15,196 43.6 19,659 56.0
1950 -------------------- 42,826 23,560 55.0 19 266 45.0
1960 -------------------------------- 53,024 32,797 61.9 20:227 38.1
1970 -------------------------------- 63 450 39 885 62.9 23,565 37.1
1973 -------------------------------- 69:337 44:653 64.4 24,684 35.6
1974 --------- 70,830 45,784 64.6 25 0AA 35.4
1975 72,523 46,867 64.6 25:6gg 35.4
Source: U.S. Department of Housing and Urban Development "HUD Statistical Yearbook, 1976" (Washington, D.C., Government Printing Office, 1977), p. 261.
See Fisch, Oscar, "Dynamics of the Housing Market," Journal of Urban Economics, October 1977.
2 Sternlieb, George and Robert W. Burchell, Re8idential Abandonment: The Tenament Landlord Revisted (New Brunswick, N.J.: Rutgers University, Center for Urban Policy Research, 1972).
Homeownership has been the beneficiary of positive attention from practically all groups within our society. As indicated by poll after poll it is the keystone of the good life for most Americans. Thus the literature on homeownership, both popular and technical, is vast.'
Large scale multifamily housing on the other hand has uniquely been undertaken by the professional developers and managers. These relatively specialized individuals are competent to deal with the increasing role of Government both directly and indirectly in its provision.4 The bulk of these interventionary mechanisms is by no means the province of the Federal Government. While conventional tract developers bemoan the increasing load of locally mandated requirements; minimum lot size, subdivision controls and the like; acquisition of land for individual dwellings, though frequently costly, has been far more popularly accepted than equivalent provision for multifamily units. Even at luxury rent levels the latter are frequently barred from suburban locations Multifamily housing is viewed as the city, as crowding, as changing the nature of the suburban-exurban setting of the homeowner. When the development incorporates low-income housing-the situation is even further accentuated.8 The post-World War II shift of the basic shelter ethic of the United States from rental accommodations to those offering home ownership has reinforced this feeling. The resulting stress, as reflected in the escalation of the costs of land appropriately zoned for multifamily housing in desirable locations has been enormous.7
In the last several years the real housing buying power of American households has been substantially reduced. Inflation in land, in building cost and, most of all, operating elements have begun to far outstrip post-tax incomes.8 Despite this situation, Americans continue to pursue onefamily homeownership with increased vigor. Some observers believe this represents the desperation of buyers who desire homeownership at any price today, fearing that tomorrow it will be even further out of their reach. Such a condition is by definition precarious. While we may hope for the abatement of housing cost inflation, the changing demographic characteristics of America's population, together with a vast level of internal mobility, reinforces the belief that renewed attention to the muibjamily housing development area is past due.
SEQUENCE OF THE PAPER
This paper has as its prime target the definition of future demand for multifamily housing. In order to achieve this, the presentation will
3 See for instance: Marcuse, Peter. The Financial Attributes of Home Ownership for Low and Moderate Income Families (Washington, D.C., The Urban Institute, 1972); Struyk, Raymond J., Urban Honeownership; The Economic Determinants (Lexington, Mass., Lexington Books, 1976); Burnbaum, Howard and Rafael Weston, "Homeownership and the Wealth Position of Black and White Americans" (Cambridge, Mass.: Program on Regional and Urban Economics, Howard University, 1972); Heald, David, "The American Dream: Fact or Fiction," Real Estate Appraiser, July-August 1977.
4 U.S. Department of Housing and Urban Development. HUD Condominium/Coo peratire Study (Washington, D.C.: Government Printing Office, 1974); James, Franklin, The Return to the Central City (Washington, D.C.: The Urban Institute, 1978); Sturrnieb, George and Kristina Ford, Loft Conversion in New York City (New Brunswick, N.J.: Rutgers University, Center for Urban Policy Research, 1978). See Williams, Norman, Jr. and Thomas Norman, "Exclusionary Land-Use Controls: The Case of Northeastern New Jersey" Syracuse Law Review, Vol. 22, 1971, "Suburban Snobbery" The New Republic, June 26, 1971.
6 Masotti, Louis H. and Jeffrey K. Hadden, The Urbanization of the Suburbs (Beverly Hills, California: Sage Publications, Inc., 1973); Sternlieb, George, The Garden Apartment Deoelopment, A Municipal Cost. Rerenue Analysis (New Brunswick, N.J.: Bureau of Economic Research, 1964). 7Williams and Norman, op. cit., Babcock, Richard F., "The Courts Enter the Land Development Marketplace," City, January/ February 1971; Mandelker, Daniel R.," Controlling Land Values in Areas of Rapid Urban Expansion," University of California at Los Angeles, Law Review, Vol. 12, p. 734. 8George Sternlieb, et al., "The Private Sector's Role in the Provision of Reasonably Priced Housing," Federal Home Loan Bank Board Journal, spring 1976.
turn first to data on who lives in rental housing partitioned by the type of structure they occupy. By matching these characteristics to future projections of America's population (which is undertaken in the section that follows) a preliminary profile of future rental demand 'Vmerges. This is refined by unit size configuration to reflect only multifamily (five units or more) rental housing. This in essence raises the question, "If all elements other than changing demographic characteristics of the population are equal, what would be the demand for multifamily rental housing?"
Clearly, however, these stable conditions rarely are maintained. Thus, the third section of the study looks at the supply elements and the factors which impact upon them. This, in turn, reflects upon some of the underlying dynamics which may weU significantly alter the demand silhouette generated in the preliminary approach. The last section of the studi concludes with a summary of basic findings together -with their implications for national housing policy.
I. WHO LIVES IN RENTAL HOUSING: A MARKET PROFILE
Practical forecasting requires a firm foundation in the present. Rental housing, particularly in larger structures, appeals to certain segments of our population. This section of the study is devoted to defining these elements. Some tenants are there by choice-others by necessity, some are loag-term residents and some are transients, either in hope or reality, on their way to one-family facilities. In any case, this type of profile when applied to future demographics provides first rough estimation as a basis for more refined subjective elements.
HOUSING AND POPULATION CHARACTERISTICS: HOUSEHOLD SIZE BY STRUCTURE
There were more than 25 million renter households in the United States as of 1975. Typically these were relatively small households in terms of the number of people within them, indeed the median figure for all renter households was 2.1. Slightly less than one-third are oneperson households, with a nearly matching proportion containing two persons. Households with five or more persons make up only one in nine of all renter units.
See Exhibit 2.
EXHIBIT 2.-RENTAL'HOUSING-HOUSEHOLD SIZE BY STRUCTURE SIZE (U.S. TOTAL, 1975) fin thousands
(U.S.) 1 2to4 5 to 9 10 to 19 20 to 49 50-plus home or Persons renter unit units units units units units trailer
Total ------------------------------------ 25,656 8,432 6,772 3,028 2,514 2,058 2,332 519
1- .............--------------8,262 1,589 2,175 1,119 1,019 936 1,279 145
2 ------------------------------------- 7,733 2,313 2,153 966 832 665 650 153
3-----------------------------4,187 1,630 1,181 443 377 261 189 106
4--------------------------------- 2,719 1,293 724 343 155 116, 113 75
5 ------------------------------------- 1,392 740 318 130 81 45 53 25
6 or more ----------------------------- 1,364 867 221 127 51 35 48 16
Median (United States) ----------------- 2. 1 2.7 2.1 1.9 1.8 1.6 -------- 2. 3
Northeast North-Central South West
Regional rental units (total) ------------------ 6,690 5,938 7,763 5,254
1 unit ------------------------------------ 982 1,918 3, 610 1,922
/t r............................- 46 98 272 103
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975)," table V-53-45 (unpub. lished).
Most renters do not live in large-scale structures.' Approximately
_Kristof, Frank S., "Urban Housing Needs Through the 1980's" (Washington, D.C.: U.S. National Commission on Urban Problems Research Report No. 10, 1968).
one-third live in facilities with oniy one unit (i.e., one-family rental units), with an additional one-fourth in structures with two to four units. In deed, if we were to limit the definition of large-scale multifamily housing to those structures with 20 or more apartments, the total of 4.3 million would make up barely one-sixth of all rental units.
In- general, household size decreases with the scale of the'structure. Thus, for structures with one-rental unit the median size is 2.7, for the five- to nine-unit structure it is 1.9, for the 20- to 49-unit structures 1.6; in structures with 50 or more units, over half of the apartments have only one person. There are less than 200,000 families with five or more 'persons in large scale structures-those with 20 or more units [a rather substantial proportion of these are probably public housing]. In sum, therefore, the most important configurations in the world of renter
-shelter are small-scale structures and these are particularly skewed toward smaller households, contrary to common perception. This latter element is particularly the case in large-scale rental structures (five units or more) which are substantially the domain of small households rather than large sized families.2
At the bottom of exhibit 2 is shown data by region for rental housing and some of its elements. There is an 'Unfounded belief that it is the Northeast which is the dominant province of rental shelter. Clearly, as shown in the data, this is not the case. The South actually has more rental housing units than the other regions shown, with the Northeast second, and the West last.
Included is further information indicating the number of rental units which are in one structure as well as mobile homes -or trailers. The resulting skew in regional allocation is evident. The Northeast has the'smallest number of one-unit rentals as well as the smallest number of mobile homes or trailers used for residence. These two groups combined, make up less than one-sixth of the total rental units in the region. In the South, on the other hand, more than half of all rental facilities' are in one-unit structures, mobile homes, or trailers. The equivalent for the North-Central and Western States is roughly one-quarter and one-third, respectively. Thus it is the Northeast which, by far, has the greatest number of multifamily rental structures. In that region structures with two or more units shelter a total of more than 511million individuals to the South'sg less than 3 million. 1 4 The other two regions are intermediate.
The (lata on number of persons per household earlier presented mask a substantial skew of sex distribution within households. Rental housing is much more the province of female-headed households, for example, than holds true of owner-occupied. units. As shown in exhibit 3, one out of six of all renter housing units in the United States is occupied by a female-headed household; in owner-occupied f acilities, the equivalent is one in nine. Less than one-quarter of these rental facilities is occupied by husband and wife plus children under
3 For comparison purposes "multifamily rental structures" are used here as structures for rent with tw a units or more. In a subsequent portion of this paper this will be narrowed to the more traditional definition of structures of 'five units or more.
4 George Sternlieb and Hughes, James W., Revitalizing the Northeast (New Brunswick, N.JT.: Rutgers University, Center for Urban Policy Research, 1978).
the ago, of 18 and, as would be guessed, the bulk of these are in smallscale structures. When examination was undertaken of female-headed households with two or more persons, there was little in the way of a regional skew.
EXHIBIT 3.-RENTAL HOUSING-HOUSEHOLD TYPE BY STRUCTURE SIZE (U.S. TOTAL, 1975) [In thousands
Total -Mobile (U.S.) 1. 2 to 4 5 to 9 10 to 19 20 to 49 50-plus home or
Household composition renter unit units units units units units trailer
1-person households ------------- -------- 8,262 1, 589 2,175 1, 119 1, 019 936 1,279 145
2-or-more person household --------------- 17, 394 6, 843 4, 597 1, 909 1, 495 1,123 1;,053 375
Husband and wife.----------11,632 4,992 2,911 1,173 913 688 693 263
(With own children underfi 18)------- (6, 069) (3, 025) (1, 473) (598) (377) (238) (197) (161)
Other male head--------------------- 1,463 471 370 184 154 143 98 43
Female head ----------------------- 4,299 1,380 1,316 552 428 291 262 70
Northeast North-Central South West
I-person households -------------------- 2, 215 2, 041 2, 226 1,780
2-plus-person households, female headed ----1, 179 922 1, 349 850
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975)," table(s) V-53-4 to V-53-7 (unpublished).
In sum, husband and wife households represent a minority of the occupants of rental housing. This is particularly the case in large-scale 'units'with barely one-third of the households in 20-or-more unit structures having this configuration.
While blacks make up 10 percent of total'American households, they occupy 1 in 6 of all rental units.' Indeed, in the South more than one-quarter of all rental units are so occupied. There is significant regional variation along this line. In the West blacks occupy 1 in 12 of the rental units, in the North-Central States 1 in 7, and in the
Northeast 1 in 6.
As shown in exhibit 4, the distribution of blacks to whites as a function of structure size has little consistent variation un til we turn to the structures of 50 or more units. Here blacks make up fully 19 percent of the occupants-this is probably due to the inclusion of large-scale public housing facilities within this structure size grouping.
EXHIBIT 4.-RENTAL HOUSING-RACE BY STRUCTURE SIZE (U.S. TOTAL, 1975) [In thousands]
(U.S.) 1 2 to 4 5 to 9 10 to 19 20 to 49 50-plus home or
Race renter unit units units units units units trailer
Total -------------------------------- 25, 656 8, 432 6, 772 3, 028 2, 514 2, 058 2, 332 519
White------------------------------ 20, 788 6, 862 5, 521 2, 411 2, 027 1, 715 1, 771 481
Black ----------------------------- 4,252 1,428 1,164 514 412 264 436 34
Other------------------------------ 616 142 88 103 75 79 124 5
Northeast North-Central South West
Regional rental units (total)------------------- 6, 690 5, 938 7, 763 5, 254
Black------------------------------------ 1,036 804 1,979 433
Other ------------------------------------- 143 69 88 315
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975)," table V-53-1 (unpublished).
5George Sternijel and Lake, Robert W., "Aging Suburbs and Black Hiomeownership," The Annul. of the American Academy of Political and Social Science (Vol. 42, November 1975),.
There are few more inclusive socioeconomic indicators in a society such as ours than education. In this context it is particularly noteworthy to view the educational level of the occupants of rental housing of all kinds in 1975 as shown in exhibit 5. For all the heads of household in rental units, the median fipre is 12.4, i.e., slightly over the high school level. This compares with 12.2 for homeowners as a group.
EXHIBIT 5.-RENTAL HOUSING-YEARS OF SCHOOL COMPLETED
(U.S.) 1 1 to4 5 to 9 10 to 19 10 to 49 50-plus home or Years of school completed renter unit units units units units units trailer
25,656 8 432 6,772 3,028 2,514 2,058 2,332 519 No 260 143 59 9 16 3 29 1
Less than 8 2,858 1,300 695 272 155 166 222 47
8 yrs ------------------------------ 2,184 765 611 205 165 129 251 53
I to 3 4 240 1,531 1,189 501 331 284 288 177
4 8:062 2,540 2,271 932 799 635 679 206
I to 3 4,051 1,155 1,033 572 525 395 309 61
4 yrs or more- 4,000 998 915 536 522 446 553 30
12.4 12.2 12.4 12.6 12.7 12.7 12.6 12.2 Northeast North-Central South West
Median education by 12.3 12.5 12.3 12.7
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975)," table 53-51 (unpublished).
There is little variation among renters in educational attainment as a function of structure category. There is some indication that, in general, a higher level of formal schooling tends to be found in the larger developments. The lowest level of educational attainment, 12.2 median years of school completed, is found in one-unit mobile home and trailer rentals.
On the bottom of the exhibit is shown median education of renters by region. This is roughly in accord with national trends. In any case it shows relatively slight variation.
Thus, large-scale rental homing currently is skewed toward small household sizes with a disproportionate number of them jemale-headed and a somewhat higher proportion of blacks than is the casefor ownership housing. It should be kept in mind, however, that households in all forms of rental units, while representing some measure of dispersion, are well within the mainstream of American households generally.
II. HOUSEHOLD FORMATION AND THE FUTURE DEMAND
FOR MULTIFAMILY HOUSING
The rough base of the demand figures to be generated for future multifamily housing demand at its initial level is a projection of population
growth by age and sex segment. This, in turn, when multiplied by headship rates 1 (the proportion of people within each age and sex group, who are heads of household) yields the number of renter households. Once this has been established such renter households are allocated to structure of various sizes.
HOUSEHOLD TYPE AND STRUCTURE SIZE
In exhibit 6 data by age of head are presented for various configurations of household in 1975. These are further partitioned into structure size categories.
EXHIBIT 6.-HOUSEHOLD TYPE AND STRUCTURE SIZE (U.S. TOTAL, 1975) [In thousands]
Household composition (U.S.) 2 to 4 5 to 9 10 to 19 20 to 49 50-plus home or
by age of head renter 1 unit units units units units units trailer
Total ------------------26, 656 8, 432 6, 772 3, 028 2, 514 2, 058 2, 332 519
One-person households ------ 8,262 1,589 2, 175 1,119 1,019 936 1,279 145
Under 65 yrs -----------5,559 1,000 1,472 850 748 682 700 106
65 yrs and over ---------2, 703 589 703 269 271 254 578 38
Two-or-more-person household -----------------17,394 6,843 4,597 1,909 1,495 1,123 1,053 375
Male head, wife, present no
nonrelatives--.............11,517 4,912 2,895 1,167 913 682 688 260
Under 25 yrs -----------2,299 811 642 279 249 146 74 97
25 to 29 yrs ------------ 2,555 1,008 698 301 237 143 103 65
30 to 34 yrs ------------ 1,416 681 327 137 104 76 57 35
35 to 44 yrs ------------ 1, 741 911 362 165 96 81 97 29
45 to 64 yrs- ---------2, 398 1,102 599 198 135 145 187 32
65 yrs and over---------1, 109 399 268 87 91 90 169 4
Other male head ------------1,578 551 386 190 154 150 102 45
Under 65 yrs -----------1,469 508 353 188 150 139 85 45
65 yrs and over--------- -109 43 33 1 4 10 17 0
Female head -------------- 4,299 1,380 1,316 552 428 291 262 70
Under 65 yrs -----------3,918 1,238 1,199 504 404 273 232 68
65 yrs and over--------- 381 142 117 48 24 19 30 2
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975)," table A-1 (unpublished),
By using the ratio of the number of households headed by an individual within each of the sets to the total population of that same category, we secure a ratio of households to population for each of the several categories. For example, in 1975 there were a total of 5,559,000 one-person households under age 65. The total population from 18 to 65 at that time was 124,880,000. Thus the ratio of household headship was 0.045 (heads of household under the age of 18,
1Bernard J. Frieden and Solomon, Arthur P., The Nation's Housing 1975-1985 (Cambridge, Mass.: Joint Center for Urban Studies, 1977).
a very small proportion of all households, have been excluded).. By then using the U.S. Department of Agriculture projections for equivalent age groups for 1980, 1990, and the year 2000, respectively, multip lied by the household headship rate as of 1975, we secure the equivalent number of renter households for each particular category. Thus, in 1980 there will be 6,085,000 one-person households under the age of 65, in 1990 6,701,000, and in the year 2000 slightly over 7 million.
By summing the total future renter households for each of the categories a first approximation of future demand is secured. From 1975_ to 1980 the increase in total renter households is approximately 2.5 million units (from 25,656,000 to 28,226,000). In the decade of the 1980's using procedures described above, an increase of an additional 2.9 million renter households. is observable. In the 10 years from 1990 to 2000,' tk-e increase slackens slightly to under 2 million households. See exhibit 7.'
C >% CO" Ul) -it -4 M r%. f" r-- r-- Ul I
ISO -4-4 "2 e
I -d' (.0 F%. Ln r.. cn
CL mm cm
CD m -01 m U' -d* 00 U* M(D&M co W) (D W rl%
4= 44 ".4 4n 00 1" m
r-z'rw CFev v! z Le C"
c m .4t 00 qt nt mr cn -4 M m
4% rl% tDA cn r-4 cn 4="
Oi llliq: Oi W00
CL C44 CM
Lai MmgknLn Ln Rr 00 4 -4
C) CNI Ln vt r -4 to Ln cc
c c 00 en CD (m 00 4000 -40)
mmwwc:l wc> m 1.4
W M LO al-t -cr -4 r, co
CO) cc b- 4,
CL m -.4 .4 ".4 C.4 '.4 W -4 to
V) CD (D 00 = gm cy) m U) -4 U') Co (D
"0 co w ul) C-i 4m Cj 4D M "
LAJ 0 00
V) c Ln m (.0 Ln -4 -tr" "M
0-0 C) -cr C14 WL0tn'4" C4.j co C14
4="-4 -4 2) C) C) (m CL
= 00 r- a (Z
w C) cc
C) U") 00 o) -cr w mto 00
00 C -4,94, wt M r m r .4, C14
"o 00 qc
(D qcr 04 m m m
C*4 CM =44 "4
uj CL C,
=2 4m Ct) a) r-4 -4 00 cn CD co -4 to
U*O 4= 0) r.% 14t = -400 LCJ
LO rl Cj 01 r X"q *r C)M to
LO I I I I C-4 W 0
L6 1 1 1 -4 0)
O.W- - W w E -a
= 0 Cc to, Ea w. r
o o,5e E r Im Owls
E E 0 '4) Ln C, in t = 00
(n CD "Z
cc1" 0 .01 4)
= = LC) Ln'n L., s L,
Cn Ln 4) CD
E CL r-L
U- 06 CL
It must be stressed that this apprLoach is at best a first approximation with a number of very specific limitations. Principal among them is the issue of future population growNth. At least through the year 1990, however, this plays a relatively small role with all but a very few of the heads of households through that year presently countable-if subjec t to attrition. Even when the projection is taken to the year 2000, the vast bulk of the heads of household similarly are presently countable.
In addition, however, there is the proportion of individuals in the various age and sex cohorts who will form householdIs. Second, within the total growth of households, will the proportion who turn to rental facilities as against ownership remain reasonably constant? Much of the housing demand since 1950 has been a function of higher headship ratesof the high proportion of individuals willing and capable of initiating new households. There is evidence that this process, particularly for large-scale facilities, has ran its coarse. Thus, it is felt that the 1975 rates are appropriate
T1HE FIRSTl APPROXIMATION OF FUTURE RENTER HOUSING NEED
The number of new renter households is not directly equivalent to the number of additional rental housing units required. The principal adIditions lie in the issues of replacement, i.e., the loss either by demolition or conversion of extant rental units, as well as the level of vacancies that must be maintained in order to provide reasonable market fluidity. Neither of these elements is without controversy in terms of appropriate proportions.
The replacement ratio is a particularly controversial one. There is a substantial shift of rental units into the ownership stream through such processes as condominium conversion, the movement of partially seasonal rental housing into full-time ownership residences and the like. Offsetting this, on the other hand, are the uncounted numbers of units that shift from ownership into the rental market.3 Examples of the latter are conversion of single large units into two or more smaller rental facilities. The last section o~f this study will turn, to some of the potentials in the conversion realm, for the moment it will be assumed that the inflow and outgo by conversion into the rental pool are relatively balanced.
THE ATTRITION PROBLEM
There is a substantial leakage from America's housing stock in the form of abandonment and demolition. In the 5 years from 1970 to 1975, the loss in structures with five or more dwelling units approximated the 2 2 percent level.4 Much of this, based on somewhat incomplete data, was concentrated in the older cities of the Northeast. The complement of population decline, and the outmigration of the more affluent, combined with some level of provision of competitive alternative housing units elsewhere 'was respons'ible, in large part, for this occurrence.
It should benoted in this context, that the overall housing unit attrition level hovers close to the 1 percent mark, thus indicating the significant stress on larger scale structures.6
3 Michael J. Swiibrast and Seldin, Maurey, Comnponents of Future Housing Demand (Washington, D.C.: National Housing Center, 1966).
4 Comparison of 1970 multifamily (5 units or more) housing units plus annual, multifamily housing starts (1970-1975) to multifamily housing units in 1975.
5 U.S. Department of Agriculture, Forest Service, Projections of Demand for Housing by Type of Unit and Region (Washington, D.C.: U.S. Government Printing Office, May 1972).
For the purposes of this study, a 2-percent future attrition level has been projected. This indicates an average longevity of 50 years for the multifamily stock. It assumes that while the unnatural attrition level of recent years may continue in the central cities, it will be somewhat blunted as a national statistical phenomenon by the increasing level of multifamily housing which is either relatively new and/or constructed in areas of substantial housing demand.
Vacancy levels are the essential lubricant of the market. There is no generally accepted definition of a "market" which is an appropriate balance between supply and demand. Alaska, for example, has defined a housing market with less than a 3-percent vacancy rate as being of undue rigor.6 The FHA, and at least in more stable areas the housing field as a whole, has used a 5-percent -vacancy figure as an appropriate one.7 8 The 5-percent figure is used in terms of the additional units required by future renter households in order to provide fluidity and a competitive market.
Exhibit 8 summarizes, by type of household-(1) the number of future renter households, (2) a vacancy figure which adds 5 percent to this marginal increment, and (3) a replacement figure of 2 percent. This latter has as its base the midperiod level of units (i.e., 1975, 1985, 1995) for each of the projected time periods.
For one-person households, under the age of 65, in 1980, therefore, the figures would indicate (1) a future renter household number of 527,000 (see exhibit 7 for the origin of this figure), (2) a vacancy figure of 5 percent-or an additional 26,000 units-and (3) a 2-percent replacement requirement (582,000 housing units) based on the number of renter households within this category as of the midperiod prior to project. This yields a grand total of 1,135,000. For each of the household types, and for each of the several intervals through the year 2000, the data are presented in exhibit 8. At the very bottom of the exhibit are shown the totals for all types of rental demand using this approximation procedure. Thus, for the period from 1975 to 1980, by these calculations, nearly 5.4 million additional rental units will be required. For the decade of the 1980's the equivalent figure is close to 9.4 million; for the 1990-2000 period it is over 81% million additional rental units.
It is interesting to note the impact of both the baby boom and the baby bust in the data array.9 The first is shown by the single largest
6 See Monica R. Lett, Rent Control: Concqpts, Realitic3 and MAchalniMs (New Brunswick, N.J.: Center for Urban Policy Research, 1976) p. 40.
7 This is a shade less than the current rental vacancy data. While there is some regional variation, with vacancy rates typically higher in areas with substantial construction rates (the South and West versus the North) there has been an increasing tendency to leveling across the Nation. Thus the future total vacancy rate projected here is the sum of the current rate plus the provision made for new units.
S U.S. Department of Housing and Urban Development, FHA Tcchniques of Husing Market Analyis (Washington, D.C.: U.S. Government Printing Office, August 1970). 4IDe Leeuw, Frank, The Demand for Housing: A Raitw of Cross-Sedional Eridnce (Washington, D.C.: The Urban Institute, 1971).
m Ln gn
"W"=W w CD W C cn m r.. -.4 00 C" w = r, en UM --4 P%
C> 0 _,. -*WWWM m 40..4 40
00 1.. "t w Ci -4
CD -d- U-) cn 00 r- tn cn co r UM
00 -cr en Uj Ln Cn 00 P, m co
-tt C-4 comr, ul
-cr LM tn
C"i 00 a,* ul) w -mr = CD Lo 00
4m C14 w cn W Ul) U") m .4 m
00" w 00 co tow w -W
cn r W Lncnr-cnw cn Cj m m Go
f.0 a) -:T -cr m 00 P.
Cj r.. -4 co vt
CD lqr 1-4 Cj 00
CL W-0 Ln M .*M "rl- Ln
Wuj MWW,-$" c"
76 UC= Ln Ull) mr U' u W w --* M -d*.-4 00
m ul -4 (D m W m 00
orwen C" 0000 M M W Ln
W -o:!-- N... Ul) Od
MW W-R CT
0? C= be
w cc V -!t w w r cm
cv m v-4 1-4
V) cn Co Lm qp ka P.. c"
= -tt -4 -4 19t
LO CY) to CL
(D Ln G)Ln
> cj -cr -cr
1=0 Zll Z
W 0.2 -0 a)
m 0. 0 ba
4) =LnLnLn LO
=) Cj M .4t (.0 wl; (.0 9=
06M = E
sector increment in 1975-80 l~eriod; it is in the 25-to-34 age category for male-headed households with wife present. In the next decade it is shared by the 25-to-34 and 35-to-44 year groups while in the decade of the 1990's it moves to the 45-to-64 year bracket. The baby bust, on the other hand, is shown by the two areas of decline in the decade of the 1980's and 1990's, respectively. In the first it is the male-headed household, wife present, under the age of 25-in the second, as a reflection of the age cohort advancing in time, it is the same category of household, but now aged to the 25-to-34 year mark.
It is equally important to note the striking increment in female-headed households under the age of 65 with two or more persons. This distinctive group requires more than 700,000 additional rental housing units from 1975 to 1980, over 1,300,000 in the following decade, and by the 1990's, it is the single largest of the renter household categories.
T HE SHARE OF THE RENTALHOUSING MARKET To BE SECUnED BY MULTIFAMILY UNITS
The data shown above are for all rental housing units but as shown in exhibits 9A-9C, there is substantial variations in the proportion of households of various configurations in terms of the types of rental structures in which they live. In the material which follows, incidence within -various structure categories as of 1975 by household characteristic is used to pro rate the future demand for multifamily, renter-occupied units distinct from the total universe of rental housing.
The data earlier discussed in this paper for household incidence by housing type make it possible to further refine and. secure a projection for the various configurations of structure which will be required to incorporate the rental units. Again it must be stressed that these projections depend upon 1975 equivalence, i.e., respective categories and preferences similar to those of the specific base year. The projections which are shown in exhibits 9-A, 9-B, and 9-C incorporate both the vacancy rates and replacement ratios earlier discussed.
They show that in the 5 years from 1975 to 1980, a total additional multifamily rental demand slightly in excess of 2 million units-2,084,000 or approximately 416,000 units per year. For the 10-year period from 1980 to 1990, the equivalent is 3,674,000 units-or just slightly over 367,000 per year. In the decade of the nineties, the multifamily demand (5 units or more) declines again to the 335,000 mark per year.
It is the scrappage and replacement level rather than new household demand which is the major factor. Thus, the success of efforts at multifamily structure rehabilitation or other forms of preservation becomes the crucial variable in assessing future needs.
C* P* M r.. m V-4 C" to) a
:w m -4 t;r 04 m
C" 4000 ul) -tr CD w 0 LM
E m gn O.N. CM m4mmul)c" IM-4 CD(.D=
'0 LO 40 -I;t M CJ V-4 -4 C14 00 m
cn (A -tr CJ w r-, m w
r, .0 -cr cli CY) CN CD ---4 m
Go m C 1.0 Q,) 00 to u') m c4i 4
V-4 Cl C14 "r
-d' to 00 m Lm -e 00
=M mcn -"Oo m -;T m r
C%4 1-4 -4 -e CJ 4
c:p mt co (D Ln ;t CM 00 r- cp 4m -tt a w lj 14-4 co 4= r- 4m CD 00 cli co
C.0 m r-4 M 1-4 ".4 04
bow U,).RrLn.OLL,) Mm %0-4-4
c (n cc CO --I m Lo -4 4D m m cn
oc 1 -4 .o cn -0, c,4
1 =,.- I E
:4D-- I c
W to:R ICE
= 1 C-4 .::r lzr -& D -6 4) 140 ca 0
-c-i-gr4o 0 0
0 0 0 0 7v co"o
4) 0 0.= C Ln Ln U-) W) Ln.2 U.)
0 V)LI m cc:
V 00 -it -4 m w w w C rn m
-c- to M-4 -cr C-4 cli r
SO: E C'a
c- "Ln cnkof.OC2 1-- moo M :4=
tr r 60 I,. U") rl IC4
P. 40 -qqr M -4 W in "t M LO
10 CV7 04 M 1.0 M wr qr cn Les
cz cm M Lo CD Cj Cj 00 tv CD
W" qr r-- cn -tr &n (n ..4 =
C> cn cv"mmCo UICD cnc""
M" 0-4 00 C-4 tn -tr co P. Cj Q
CD -cs, r-*OM4=M "Ln =r"Ln
Lnm :t qr
2w 4D C: Ln g m w c. en m Ch
mr 00 C>-4mw P'" "ra-It
M" rl rl. C*4 -4
cn C> U-1 mw..tr-= Mr-4 WLnr
C)to -;r cn 14 qw to
Ln = %_,
t4m w 00 4D C4 (= 00 4W C: C:j 40 4=
= "00 CnLnA-.r4= =Ln Rt 00 W M
o 0 Cj w to W) Ul cn Ln r-4
cd 4 1,45
CD- Q),DL,; 0
f 0 Lo Ln Lo Ln =Ln
a, &6 CD -;T q o Q>'a =;D"a
0-4 in -4 UD c 00
W-W wr -M
IM "C4 q -qcms
CM Go C-4 w wit 40 0) wl CD -tt Go
C*4 M CD
"t .a Ul M CD 0 rl e4 C"i 00 en tn
W 4= 1-4 LO -cr qr
CD Ln" "M C.,j
cl) M -c:r 00 CD -:r Ln LO Ln Cf) cn
cc wr cv).-Icsj D-tr r "" "tn-q
en" 1-4 Lc) W -4 Ln 00
to Cl M Cf) CD Ln M I-. -tr
4= LO C-4 M M
E cL*0 -Z C6
ow ul E
F-10 E Ln
Cj -cr CD I',
to Itt C14 r M.-4 I"
cl I( Do 10, 4 IN In
&n VO W
cj cr,4r qr to (D
w=="cn-w4c -=w fALn = 1- 0 cc E
III. SUPPLY AND DEMAND FACTORS IN
THE PHYSICAL DEMANDS OF CONSTRUCTION
Neither the projected number of gross rental housing units nor the proportion of them allocated to multifamily structures is such as to cause alarm over the delivery capacity of the American building industry. The target fig ires shown, even extending them substantially,
are well within our grasp.
America's delivery capacity in multifamily housing is evidenced
by its experience from 1969 through 1977 as shown in exhibit 10. The peak year was the 917,000 units in such configurations, started in 1972. This undoubtedly reflects the substantial impetus of section 236 funding.' The trough is in the recession of 1975 with starts at 208,000 during this year decreasing to barely 20 percent of the peak level.2 For the 7 years from 1969 to 1975, we started more than 4 million
multifamily units or an average slightly in excess of 600,000 per year.
EXHIBIT 10.-TOTAL (PRIVATE AND PUBLIC) MULTIFAMILY (5 UNITS OR MORE) HOUSING UNIT STARTS, 1969-77
Total (private Federally subsidized
multifamily Total (5 units Percent of
or more) or more) total starts
1969 -------------------------------------------------- 590, 100
1970 -------------------------------------------------- 558,000
1971 -------------------------------------------------- 798,500
1972 -------------------------------------------------- 917,000 11,045,370 ---------------1973 ------------------------------------------------- 800, 300
1974 -------------------------------------------------- 386, 800
1975 -------------------------------------------------- 208,100
Total (1969-75) -------------------------------------- 4, 258, 800 1, 045, 370 24. 5
1976 -------------------------------------------------- 292,200 54,340 18.5
1977 -------------------------------------------------- 415,200 179,430 43.2
I Cumulative, 1969-75.
Note: Limitations of data do not permit precise count of the number of multifamily units in nonrental forms. Even using the gross numbers of cooperative and condominium construction, shown in exhibit 13, would not alter the conclusions, however. The approximate number of multifamily (5 units or more) condominium starts was: 1974, 104,000; 1975, 26,000; 1976, 38,000; 1977, 57,000.
Source: U.S. Department of Commerce, Bureau of Census, "Construction Reports (C20-78-1)--Housing Starts," January 1978; 10th Annual Report in National Housing Goals, January 1978, Washington, D.C., GPO, 1978, p. 40.
BARRIERS TO MEETING DEMAND
High Rate of Government Subsidy and Inflation
In more recent years, a slow but increasingly significant postrecession upsurge has started with the 292,000 starts of 1976 followed by 415,000 in 1977 and somewhere on the order of 550,000 to 600,000 anticipated for 1978. It should be rioted here that the new vigor is substantially a
I Houing and Cammunity Dr(oprnen Ad of 1968, "Section 236 Interest Subsidy Program."
2 See George Sternlieb and Jughts, James W., Curraid Population Trends (New Brunswick, N.J.: Rutgers University, Center for lUran Policy Research, 197).
reflection of an increased level of explicit government subsidy. Practically all of the increase from 1976 to 1977, for example, was the result of Federal subsidization .'
The need for this level of Federal input is a result of the growing imbalance between rent paying capacity and the rents required for unaided new multifamily units. In exhibit 11 is shown data on the evolving picture of new major structure construction torrether with an equivalent time series for housing costs generally as well as the changes in the cost of "all items." It is the first of these which has risen most dramatically. The cost of new major structures has gone up by nearly half [49.9 percent] from 1972 to 1977. During the same period the "all items" index rose 41.3 percent, general housing costs 42.5 percent.4
EXHIBIT 11.-INFLATIONARY INDEXES PERTAINING TO THE HOUSING INDUSTRY, 1972-77
Consumer Price Index (CPI) Boeckh Indexcost of new
All items Housing construction
.(percent (percent (percent
All items increase) increase) increase) 1
1972 ------------------------------------------------------- 125.3 129.2 100.0
1973 ------------------------------------------------------- 133.1 135.0 105.9
1974 ------------------------------------------------------- 147.7 150.6 115.8
1975 ------------------------------------------------------- 161.2 166.8 127.2
1976 ------------------------------------------------------- 170.5 177.2 131.3
1977------------------------------------------------------- 21771 '184.1 149.9
1 Apartments, hotels, and office buildings.
2 41.3 percent increase over 1972.
a 42.5 percent increase over 1972.
Note: GPI-1967 equals 100; Boeckh Index-1972 equals 100. Source: U.S. Department of Housing and Urban Development "1976 Statistical Vearbook-HUD," p. 258; U.S. Department of Commerce, "Construction Review," November 1977, p. 47.
It is most striking to note that this level of inflation was undented by the massive recession in multifamily starts in 1975 which saw a precipitous downturn of nearly 80 percent. While perhaps a longer period of recession might curb the cost elements, the secondary ramifications (do not make this an appealing prospect. Land Zoning and High Interest Rates
Principal among the cost elements involved are those of land and money. The problems of zoning and the increasing flight from the dollar into land speculation have made sites which are both suitable and legally available for multifamily housing a relative scarce commodity in areas of high demand.'
The consequence in terms of the impact on the rent level requirements need little amplification. While a number of States have made slight overtures toward a broadening of land use elements, these have been relatively minor. Many of the issues of zoning for multifamily housing, therefore, have ended up in the courts, a procedure which is both costly, time consuming, and scarcely genleralizable.
Obtaining land zoning through the courts is both uncertain and very expensive. The construction of multifamily housing, therefore, has
3 See also Donald E. Priest, "The Unchart ed Trend Toward Increased Public-Private Cooperationift Housing D evelopment. "Ainerican Real J-.irite and Urban Economic8 Associationi Journal summere r 1977); Ira G. Kawaller "The Role of Federal Sub~sidies in the Construction of Multi-family Homes" (paper presented at Southern Economic Associ-ation Convention, November 1977).
4 U. S. Department of Commerce, Bureau of Domestic Conmmerce, Construction Review, December volume 1967-1977.
6 Williams and Norman, op. cit.
become a highly riskful venture. It has not been aided by a variety of *environmental protection requirements, many of which have been mlsuisedl as a second~ line of defense by local authorities in order to avoid high density housing
The Federal entry into land use regulation has been miniscule. While there have been occasional urges by A-96 Review Agencies to take the lead in this regardl, there has been little backing when thes.-e middle level review groups come under fire from local communities and citizenry.
The cost of money, a commodity whose cheapness in the United States made our housing the envy of the world, h as been rising across the board. It should be noted, moreover, that despite this overall increase, the financing of multifamily housing is not considered a desirable venture on the part of major sophisticated lenders."
At this writing, insurance companies are actively competing for shopping center properties. Indeed, the capitalization rate on some of the very best of these has been driven down -to the V,2 percent level. At the same time, the proportion of total loans by such companies in multifamily housing has declined-and the rates of return required are escalating rapidly. Current capitalization rates, for example, on prime apartment house construction are currently reaching the 10 percent level.7 Much of this stems from the operating imbalance endemic to this form of housing.'
Operating Costs Escalation and the Fear of Rent Control
The increased pressures on operating costs of multifamily housing are far from- unique. They ai'e felt in a broad spectrum of American life and industry. Within the homeownership domain, however, they can be at least partially offset by a combination of a do-it-yourself and don't-do-it-at-all approach. The homeowner can absorb some of the increases by doing more work within his residence than would otherwise be the case and simply not think of it as an expense-or ,otherwise reduce the levels of maintenance or resident comfort, i.e., if one's fuel bill is too high, decrease the heat level. In industry as well as shopping center operation, there have been substantial efforts to utilize a variety of capital intensive investments to reduce operating costs. Multifamily housing has not yielded to equivalent procedures. Operating standards are largely set by a combination of the market and local housing codes. There has been little in the way of innovational Jrontend investment to replace current practices of building operations.9
At most, we see a drive toward reducing or transferring the burden of costs through separately loading operating costs. Examples of the latter include the decentralization of air-conditioning, making it incumbent upon each of the tenants to pay for his own usage, separate metering for utilities, et cetera. These provide automatic pass-alongsbut, if anything, increase the gross rents since, in general, such provision is less efficient than centralized equivalents.
5 See Michael A. Stegman, "Multifamily Distress and the Conservation of Older Neighborhoods: A Cause for National Concern" (Washington. D.C.: U.S. Department of Housing and Urban Development, Office of Policy Development and Research, February 1978).
7Information secured from Goldman-Sachs investment research, spring 1978.
f For further elaboration, see addenda.
I See Stephen R. Seidel, Go-trvnui Rqndcition and Housing Costs (New Brunswick, N.J.: Rutgers University, Center for Urban Policy Research, 1978).
Since 1970 there has been a steady rise in rent to income ratios. In part, this resulted from a shift of the more affluent renters into homeownership; in part, however, it is a reflection of the increased rents required to meet operating costs.
In 1970, for all rental housing, gross rent as a percentage of income. w~as 20 percent. By 1875, it had risen to 23 percent. In that year rentincome ratios were highest in large-scale multifamily housing, with themedian at approximately the 24 percent level, as shown in exhibit. 12. While the cost squeeze is far from unique to rental housing (in-creases in costs of homeownership have been higher than tho-se'of' rental housing) homeownership costs have been offset by the capital. gains resulting from inflation. The real costs of homeownership, at least for the more fortunate* occupant, have been decreased by thie increases in housing value. Thus homeownership has been sustained],. despitee the cost-income squeeze in this arena, by its speculative nature; by capital value increments which have abated the pressures. of annual carrying costs.
EXHIBIT 12.-RENTAL HOUSING-GROSS RENT AS A PERCENT OF INCOME (1975)
(U.S.) 1 I to 4 5 to 9 10 to 19 20to49 50-plus home orRent as a percent of income renter unit units units units units units trailer
Total --------------------------- 24, 959 7, 736 6,772 3, 028 2, 514 2, 058 2, 332 519
Less than 10 percent-------------------- 1, 710 629 484 184 137 122 125 30
10-14 percent -------------------------- 3,599 1,215 987 426 350 266 313 41
15-19 percent -------------------------- 4, 095 1, 261 1, 114 490 466 321 377 66
20-24 percent -------------------------- 3, 572 1, 018 900 454 406 304 431 .59'
25-34 percent -------------------------- 3, 990 1, 076 1,121 547 447 365 354 g0
35 percent or more ---------------------- 6, 556 1, 668 1, 935 867 660 614 680 132,
Not computed -------------------------- 1,437 869 231 61 49 66 51 i11
Median -------------------------------- 23 22 24 24 23 25 24 26
Northeast North Central South West
Gross rent as a percent of income by region
(median)--------------------------------- 24.0 23. 0 22.0 24.0
Source: U.S. Department of Commerce, Bureau of Census, "Annual Housing Survey (1975), "table 0-53-53 (unpublished)..
The same does not hold true for rental housing. There is no equity buildup, the hard-pressed con sumer simply finds a larger share of his or her 'income going for rents.
The imbalance, in a variety of areas, has brought either the reality or the threat of rent control. In turn, this has had a very strong chastening influence upon lender and builder willingness to become involved in the multifamily rental housing industry except under the most favorable circumstances.'10 These largely revolve around luxury construction, the possibilities of condominium construction, and/oressentially a bailout Government mortgage. In this last case, the
builder's profit comes not from operating the structure, but rather from its construction. The price, in turn, has been a massive wave of Government guaranteed multifamily mortgages which are in deep difficulty. As of late 1977, approximately 1 in 5 of the nearly 7,000 projects currently 'under Federal subsidy in the United States were
10 Emanuel Tobier, et al., Mortgage Financing arnd Housing Markets in New York State: A Prelirn inary Report (Albany, New York: New York State Legislature, May 1977).
either in a state of mortgage default or assignment, or the projects themselves have been acquired by HUD through foreclosure."
In the privately financed areas, the situation is not nearly as serious-but there is some indication of an increased level of mortgage delinquency. The drive toward condominium conversion bears strong
witness to the reluctance of private operators to continue despite preferential tax legislation.
Cooperative and Condominium Conversion
The growth of the condominium and cooperative form of housing, particularly the former, is shown in exhibit 13. In 1970, there were only 85,000 condominium units in the United States. In the 5 years from 1970 to 1975, more than 1 million new units were constructed. In addition, there were 100,000 conversions. While the pattern in cooperatives clearly shows a preference for conversions, it also indicates new construction equal to 20 percent of the base in the 5 years from 1970 to 1975. By 1975, condominiums made up 1.85 percent of all occupied housing units of all configurations and cooperatives an additional 0.65 percent. While many of these were in townhouse configurations, as shown in the exhibit, a substantial proportion were high-rise and garden configurations, i.e., within the multifamily domain.2
EXHIBIT 13.-CONDOMINIUM AND COOPERATIVE HOUSING STOCK (U.S. TOTAL) [In thousands]
New of all
Existing, construction, Conversion, Removals, occupied
1970 1970-75 1970-75 1970-75 Inventory units, 1975
Condominiums --------- 85 1,078 100 11 1,252 1.85
Cooperatives ----------- 351 70 25 7 439 .65
DISTRIBUTION OF CONDOMINIUMS BY STRUCTURE TYPE [In percent]
NorthNortheast Central South West United States
Townhouse --------------------------- 55 35 40 55 45
Garden ------------------------------ 35 40 40 40 40
Highrise ----------------------------- 10 25 20 5 15
Total -------------------------- 100 100 100 100 100
Source: U.S. Department of Housing and Urban Development "HUD Condominium/Cooperative Study, Vol. 1National Evaluation," p. 111-2, 1974; 111-25.
Will the condominium or co-op replace the necessity for additional rental units? Certainly the preference for ownership, strongly supported both by inflationary considerations and the tax code, is
Broader public policy is caught on the horns of a dilemma: on the one hand, resident ownership has definitively been linked with positive conditions such as the maintenance and care of structures. On the other hand, given the mobility required by a technological society, one would
11 Richard D. Baron, Beverly B. Huchman and Robert KColodny, Preservinq D1.D. Assisted Miltifarnily Hausinq: An Affirmative Role for the Area Office (Working Submission to the U.S. Department of Housing and U7rban Development, November 1977); see also HUD, Budget, Justification for 1979, Part 1, G2. 12 For elaboration on this issue, see addenda.
7Iuze to view the potential shift toward the relative immobility of homeownership with some measure of concern. Rental housing for those without the required downpayment for home acquisition, for those in transit to more permanent location, and for those individuals who require the delegation of all the managerial function of ownership to others, should not pass from the scene--and, given its key nature and scale of its incidence, undoubtedly it will not.
The need for multifamily housing is a function of both household formation, and the share of those households who will find this type of shelter most appropriate to their needs and resources. Given a continuance of the 1975 market share by household type, wNe can forecast (after appropriate allowance for vacancy and demolition), a demand for multifamily housing well within our production capacity. Indeed, the production an ticipated for 1978 is nearly 1 ',2 times the annual requirement for 1975-80, slightly more than that ratio for the years in the decade of the 1980's and nearly twice as much as required in the 1990's. The "housing Problem" has shifted from, the provision of gross number of units-to a struggle for controlling their cost.
However, we caution that the aforementioned problems-high level of Government subsidy, high rate of inflation, high interest rates, escalating operating costs, the fear of rent control and condominium conversionsare indeed serious and pose potential problems to the delivery of new multifamily units.
We have a significant need for increased provision of multifamily rental housing. We have the experience, the skilled manpower and the competence to construct it. We have yet to attack the issues of the costs and with them the adequacy of existing housing programs to deal with the problem.
The Federal efforts to decrease housing costs have largely revolved around cheapening the cost of money through subsidizing interest rates and/or extending the longevity of mortgages and thus reducing amortization. Despite efforts in this regard, the imbalance between rent and incomes has continued.
Under HUD section 8, we have approached the problem differently, in essence grouping all of the subsidy mechanisms into a rent allowance."' By making the manifold stream of subsidies explicit and amass-ing them into one figure, we may have created a politically self-defeating pro gram.' The large scale of the annual subsidies involved raises some concern as to their potential longevity. We still have not come. to grips with the basic cost issues in multifamily shelter, neither in land acquisition, construction, nor most significantly, in operation. Ink this last area; while there was an aborted attempt by HUD to develop a national training program and research activity, it is noted for its, lack of accomplishment.
The gap between our production capacity and future need provides, an opportunity to refine our approaches to providing new housingr as, well as offering a. significant potential for additional upgrading of existing housing-ncreasing the demolition ratio.
13 Housing and Commnun ity Development Act of 1974, "Section 8 Housing Assistance Payment Program." 14 For elaboration on this issue, see addenda.
National data can be misleading. A vacancy in New York City does not provide housing opportunity for the family which has moved to Houston. Much of the challenge of the future must involve a selective pruning of archaic and faulty multifamily structures from areas of declining need, while maintaining an appropriate fiscal maintenance balance for the remainder. At the game time, however, the responsibility for developing new facilities for regions of high growth will be substantial. The increased trend toward government financing of multifamily housing indicates a weakness in their market viability-an inadequate balance between construction costs, financing, operating costs and rents. However, there is strong reason to believe that unless there is a significant shift out of one-family housing, a ver strong and critical review should be required of major incentive programs to increase the flow of multifamily units in any su bstantial measure in excess of the target figures presented here.
We have been coping with housing production in a relatively shortterm frame of referenc-e. Even the 10-year projections of a decade ago, and the famous 26 million unit figure which resulted, failed to grasp the longevity of housing.'5
Large-scale shelter structures make sense only with long-term utility. The apartment house built today must have utility for a minimum of 35 to 40 years, otherwise its real "costs" both to its developer and to the Nation as a whole may be completely inordinate. In turn, this basic arithmetic indicates the necessity for market projections certainly through the year 2000 and beyond.
15 Housing andi Urban Development Act of 1968.
I. Ttm~ COMPETITIVE, FINANCIAL DISADVANTAGE OF MULTIFAMILY HOUSING
In our paper we have pointed to the inbalance between current capitalization rates for multifamily housing (in most areas of the country now in excess of 10 percent) versus the equivalent for shopping centers with recently consummated sales at the 7/12 percent mark. They are both forms of income property and historically have partaken of the same financial markets. They both represent avenues for highly leveraged investment through the use of long-term debt financing. Why the difference?
The variation lies primarily in the relative ease with which the shopping center leasing format adjusts to inflation. Typical leases involve a base rent charge with a percentage override past certain minimal sales volume. Assume that the physical volume of goods-the number of units and the quality thereof remains constant. Strictly ,as a function of inflation, once the minimal rent threshold is reached, the owner of a shopping center will receive additional rents as a percentage of the inflation in total dollar sales. The escalators are automatic, they do not require renegotiation but rather take place immediately.
While the depreciation aspects of investments in shopping centers have been somewhat altered over time through chang'nes in the tax code, with additional limitations periodically under discussion, in a world troubled by inflation the shopping center looks relatively safeguarded. Thus, not only is the owner's equity relatively secure against the erosion of the dollar, but so is the collateral base of the mortgage holder.
In addition, there has been a rapid evolution of the variety of financial mechanisms used to underwrite such facilities with participation loans increasingly prevalent. A typical financing currently involves not merely a fixedl yield mortgage but also some measure of upward flexibility, either some equity participation, a potential percentage rent override on sales past a certain level, or the equivalent. Leases with individual store tenants typically call for an instant passthrough of increased costs such as taxes. The services provided by the shopping center are precisely enumerated as are the requirements of the tenantry.
The contrast with the multifamily market is evident. Rents nationally have tend ed to lag the overall consumer price index, and indeed, have even more substantially lagged the costs of homeownership. Despite this there is substantial consumer resistance to rent increases commensurate with inflation. Time lags through more or less longterm leasing arrangements are not uncommonly built into the contractual arrangement.
More formidable, however, is the fear, either real or latent, of rent
-controls. These have tended to put a damper on increases. In a study,
for example, conducted by Goldman-Sachs (the investment banking firm), the finding was made that rents generally would have to increase on the order of 20 to 25 percent before multifamily housing could be viewed as an appropriate investment vehicle for their clients.
In a sense, some multifamily housing has been the victim, at least in pa t, of a variety of federally or locally supported mortgage cost lowering schemes of the past. These have permitted initial financing at relatively lower interest charges than would have been required by free market rent levels. However, when mortgages are recast in order for owners to recapture equity, subsidized mortgages may not be available. An enormous ga then opens up between the capacity of the current rent levels to carry free market debt service requirements. For example, assume a building is financed under a below market rate mortgage at the 6 percent level, with the mortgage five times the rent roll of the building. In the course of time, the mortgage is paid down, and the owner wishes to recapture his equity by rolling over the indenture, i.e., refinancing it. Current markJ interest rates are at the 10 percent interest level. Again we will calculate the mortgage as five times the rent roll.
The difference in interest charges, very roughly calculated, is the 4 percent rate difference multiplied by five times the rent rou or 20 percent of the annual rent. While this illustration overstates the caserents, for example, may well have gone up since the initial mortgage was taken out-it illustrates the problem of moving from a subsidized market to a nonsubsidized one. The level of equity buildup in subsidized, rent limited structures, thus may be vastly overstated if just the level of -amortization is viewed.
In the case of the shopDine center, the disjuncture on refinancing is much less. While influ'en&ed by the overall increase in interest charges, there are very few that were the beneficiaries of subsidization.
The situation is further complicated by the archaic nature of operatin. ,r -patterns in multifamily housing. There has been little in the way oft technological innovation, little in the way of labor saving devices or organizational formats which will produce real savings in operating costs. The historic capacity of the middle class to live in multifamily housing without subsidization was in very large part a tribute to the availability of inexpensive labor, of janitors and superintendents paid trivial sums of money plus perhaps a marginal basement apartment in return for a 60-hour workweek. The rapid unionization of this sector in some areas, the disappearance in all areas of equivalent adequate labor, imposes a very substantial operational stress. It has resulted with dissatisfaction toward the level of services, on the part of tenants on the one hand and/or increased operating costs on the other.
The rent levels in multifamily house are also limited by the tax benefits and investment opportunities available through alternative forms of housing. There is substantial evidence of a cream skimming procedure-a shift of more affluent tenantry over time to the onefamily market. While this may be attuned to national objectives, it leaves the remnant tenant pool much, more limited in rent paying capacity (and for that matter, will).
It should be noted, in this latter context, that while a variety of Government aid programs for renters triggefin at 25 percent of income(less appropriate allowances for scale of family etceteraa, there is.
substantial evidence that occupants of one-family housing are paying substantially higher proportions of their incomes for this privilegethus indicating relative preference in the marketplace for the latter shelter format.
In sum, therefore, sophisticated investors view the multifamily structure, except under unique circumstances and unique locations, as a relatively risk ful, noninflation proof investment.
We have had substantial involvement in interviewig on an off-. the-record basis, both major mutual savings banks and insurance companies which were once major investors in this form. Suffice it to say that, in general, it is now only of marginal interest.
II. THE, FUTURE OF. SECTION 8
The history of section 8 and the stipulations (subsequently generally violated) of the original enabling legislation and administrative requirements illustrates the frustrations, limitations, and learning about housing of a whole generation of programs. The legislation was an effort to take the Federal Government out of the real estate business-out of the production subsidy business, out of the locational problems and move rather to a revival of the low income marke t through direct subsidies to consumers. Its ideal was mixed income housing and the stimulation of construction and rehabilitation through the strengthening of demand. The increasing reality is one in which all or nearly all the tenants in a building are under section 8, of a constellation of aid programs used concurrently, with section 8 essentially sitting uneasily on top of them. And this is in direct contravention of its original intent as a replacement of such programs. It is one in which section 8 is increasingly used as a takeout mechanism for poorly conceived and/or financed governmental housing efforts of the past. In this latter regard, it is being utilized to relieve the fiscal pressures of FHA and State housing finance agency projects which would otherwise require refinancing.
Thus in a significant measure section 8 merely involves the propping up of older forms of Government subsidization by new forms of Government assistance rather than their replacement.
The program is enormously costly. At a time when median rental levels in New York City, for example, hover at around the $200 per month mark, the f air market rents for new construction and substantial' rehabilitation in elevator buildings within that city range from $491 for an efficiency apartment to $873 for a four-bedroom unit. For extant housing, the equivalent figures are $223 and $390. There is increasing evidence that the nominal maximums, particularly for new construction, become the minimums. There is additional evidence that there is much questionable rehabilitation being offered-and subsidized-by the program.
Certainly some of these programs are the results of administrative difficulties to be encountered by any new, complex approach to an area as varied as America's housing. However, the track record now is long enough to raise serious question.
At a minimum it will require much more in the way of supervision. We would sug 'gest further that the unit costs are so high-the number of individuals covered by the programs so very large, as to limit its extent inate future. It does nothing to attack the basic operating cost problem, nor
does it provide adequate stimulus for operational and/or construction ejiciencies. The program conceptually is commendable-operationally we would view it as questionable.
-111. FUTURE ADDITIONS OR REDUCTIONS OF THE RENTAL HOUSING
STOCK KTHROUGH CONDOMINIUM CONVERSION OR SUBDIVISION OF
The- rental housing industry of the United States gives substantial indication of moving away from its unique operating pattern to one
-much closer to the European-particularly the French version; the condominium or co-op. In the basic paper. we have cited the (lata which is available on condominium conversion. There is great variation in the -rate at which this is taking place within the United States
-as a whole-and within the economic categories of tenantry as well. There is no question that for the more affluent members of our society, given current tax laws and the inability to pass through that portion ,of rents which go for local property taxes (an issue which is currently being raised by New York State) the condominium-co-op format has increasing post-tax virtues. These are further compounded by the possibilities it offers as a possible haven for inflation-fleeing dollar investments. We would suggest, however, that given the present income levels of the Nation's renters, the level of conversion, unless aided by some form of Government financing, will be relatively slow
In terms of new construction of multifamily units, the pattern is much more 'forcefully toward condominiums. It should be noted that the data in these areas must be viewed with some measure of trepidation. Studies that we have undertaken in Florida, for example, indicate that a number of nominal condominium units, depending upon the vagaries of the market and specific ownership patterns are available for rent. The flow between these various forms of tenure is quite abrupt and probably at least in part, avoids the Census count net. To the degrree therefore, that condo co-op replaces straight rental housing, there will be a decrease in the number of such units available.
The other side of the ledg-er, however, is the conversion of one-family private homes into two or more units. Though it has attracted much less attention, it may be equally forceful. Much of this is undertaken outside normal, legal procedures. Indeed, in a great many jurisdictions in which it is occurring-it is Specifically illegal. We have undertaken a comparison of Census data over time which indicates for more twofamily homes than can be accounted for in terms of nominal permits and starts. The answer is conversion.
The median size of household since 1970 in the United States has gone from 3.14 persons down to the 2.8 level. The increase in home,ownership costs as a percentage of income has been equally dramatic. Given these elements combined with the extraordinary number of four- and five-bedroom units, particularly in split-level configurations, yield a highly probable flow of conversion in the future.
Field studies undertaken by the Center for Urban Policy Research at Rutgers University, for example, in a classic post-World War II .suburban area-Bergen County, N.J.-very specifically indicate that this process is well underway. It should be stressed that this is not merely a phenomenon of poor or central city areas-but rather one which is also taking place in the suburbs.
Thus two conflicting elements are at work. The first, condominium conversion reducing th~e number of rental units,. the second conversion of one-family homes into two or more units, increasing their availability. A side note should be entered in this reckoning which indicates some of the fiscal pressures which are at work. One of the more popular configurations of New York City housing now-and one which dominates the unaided housing trsi h so-called "illegal three." This
is the term which is usdin official city counts to denote structures which are nominally built and licensed for two-family occupancy but Which incorporate as a matter of course a third unit. Given current real estate costs, this merely exemplifies the pressures to secure some measure of income in order to support ownership.
1HUD HAS A RESPONSIBILITY FOR ALL THE NATION S HOUSING STOCK
The bulk of HUD programing, executive focus, and research has been devoted to the low rent-income end of the housing stock. There has been inadequate focus on the long pipeline Which lies behind it, and on the enormous national investment in the general multifamily housing stock which simply cannot be replicated. Federal action, in terms of stemming the tide of abandonment and housing decay, has tended to intersect too little and too late in the process. In general, it has been based upon an inadequate comprehension of the overall dynamic. HUD simply must broaden its attention span.
2. WITHIN FHA THE LEVEL OF DATA MAINTENANCE AND COMPREHENSION ON MULTIFAMILY HOUSING IS TOTALLY INADEQUATE
While operating statements are required under a variety of FHA programs, they are rarely, if ever, audited, poorly reviewed, mnadequately administered and standardized. Second, there is no overall attempt at analysis, at developing operating cost data, trend analysis, etc. In the absence of such devices abrupt and very costly crises which may have been long in the making come as unexpected -surprises. We cannot afford the- slop piness involved. Nowhere in the United States is there adequate, impartial data on operating costs and the like for multifamily housing. We are collecting the raw elements required for such analysis but simply not closing the loop in terms of appropriate structuring and quanti-fication.'
3. TAX TREATMENT OF RENTS
The issue of the deductability by tenants of that portion of their rent which essentially flows through the landlord's hands to local jurisdictions in terms of real estate taxes must be reviewed. The situation is already coming to a head in New York and rather than a, perfunctory acceptance or Irejection of the concept a rig-orous analysis of the future role of rental housing and the issue of its tax treatment should be undertaken.
1A possible guide would be the operating data for multifamily buildings gathered under contract for New York City by the BLS.
4. THE FEDERAL ROLE IN RENT CONTROL
There should be no hesitation or reluctance to override local rent control ordinances when the latter impact the fiscal vitality of federally financed or guaranteed projects. The "on again-off again" of practice in this area has brought into question the Whole validity of HUD ,rent guidelines. It has endangered both private and public multifamily investment and, most importantly, has left the courts and the electorate without a nonpartisan yardstick.
5. CONVENTIONALLY CONSTRUCTED MULTIFAMILY HOUSING IS TOO
EXPENSIVE TO USE FOR SHORT-TERM CRISES
The scrappage rate (losses from the stock) are so high, the foreclosure and mortgage delinquency data so ominous, as to provide reason for apprehension. We have been much readier with subsidy mechanisms to launch multifamily housing, i.e., through mortgage
-subsidies and the like, than we have been to grasp the full-life cycle ,and the issues of refinancing.
There is currently substantial pressure to broaden out those provisions of section 223 which permit refinancing for rehabilitation of multifamily housing. Given the relative weakness of the market this may well end up with Uncle Sam becoming the owner of structures which come under the program by default of owners. The latter will liquidate their investment at nominal face values which simply.overstate their market worth. The principle of supporting conversion is ,essential. Its operating mechanisms, however, require much more attention.
16. THERE IS AN ENORMOUS NEED FOR OPTIMIZATION OF MANAGEMENT
AND OPERATING PROCEDURES IN MULTIFAMILY HOUSING
The Federal track record in this sphere is notable by its absence. While early experience in attempting to support an operational management research activity has been sadly disappointing, reexamination of the entire area is called for. The payoffs could be most considerable. Considering the fact that government at all levels is de facto the largest single owner of multifamily housing-and there
-is some indication if anything, the stock in its possession will increase there is both a broad as well as a parochial necessity for such programing.
7. MOST IMPORTANTLY OF ALL, IN THE LIGHT OF THE INFLATIONARY
BONANZA THAT HAS BEEN ATTACHED TO HOMEOWNERSHIP, THE DEMAND FOR MULTIFAMILY HOUSING AS INDICATED IN THE BASE
PAPER WILL NOT INCREASE
Estimates given current levels of market penetration by household characteristics multiplied by the numbers of households forecast in each of the several sectors considered, indicate that substantial conservatism must be the order of the day.
This is a very harsh reality particularly in terms of immediate need. There is a contraction in tfie availability of such facilities. But we are not dealing -with a transient good. The only way such structures make -sense is ln terms of a long and useful lifetime. The costs of overoptimism in this sphere could be enormous.
UNIVERSITY OF FLORIDA 3 1262 09118 5867