committee report which contained the recommendation. A basic assumption of the recommendation is that mortgage credit will be required to produce 16 million permanent nonfarm units during the decade of the sixties. A related condition contained in the report of the Senate subcommittee was that

for the first 5 years, new construction should average 1,400,000 units a year and for the last 5 years it should average 1,800,000 units a year, probably reaching 2 million units a year by the end of the decade.

The pattern of starts developed on the basis of the latter condition is in line with the pattern of annual net increases of nonfarm households indicated by Bureau of the Census' projections of nonfarm households.

Underlying the 1961-70 figures on new residential construction in the Senate subcommittee report is an assumption of "a high level of economic activity and an aggressive Government policy to eliminate slums and blight." In order to reflect the assumption of a high level of economic activity, use was made of a projection of the American economy in 1970, by the National Planning Association, which assumes an average annual growth rate of 4.2 percent. This rate is in line with a number of other projections of economic growth made in recent years.

Two other major assumptions were adopted in addition to those contained in the Senate subcommittee report. The first was that there would be a balanced Federal cash budget, an objective that presumably will be pursued by any administration. The second was stable prices of all goods and services in the gross national product. The latter assumption was adopted because it would presumably be an objective of governmental economic policies, and by having all dollar projections in terms of constant 1959 dollars, it avoids further complications in an analysis that is already complex.

In addition, it was assumed that the structure of interest rates and the relative distribution of savings would be about as in the fifties. Various working assumptions that had to be made, such as the average house prices, mortgage amounts, mortgage repayment ratios, and lender investment policies were based on trends indicated by experience during the fifties. Therefore, the resultant projections are dependent upon the assumptions made and are not predictions of mortgage credit requirements and of the availability of funds to meet those requirements. (Any 1960 estimates that are shown also have been based largely on the adopted assumptions since only partial 1960 data were available when the estimates were made and 1959 was generally used as a benchmark year.)

The projections which have been derived provide only a hypothetical model of investment requirements and of the flow of funds that might be available to meet those requirements on the basis of the assumptions that have been adopted. The hypothetical model is a useful frame of reference, however, for the concluding analysis of the study.

Mortgage credit requirements

Annual residential mortgage credit requirements were estimated, first, on a gross requirements basis. Conceptually, this includes all mortgage loans on new and existing homes that would be made in a year. Then, estimated mortgage repayments were deducted to derive annual net residential mortgage credit requirements.

Based on observation of recent history, a judgment (more political than economic) was made that public housing would account for 25,000 of the new permanent nonfarm units to be started each year. Also, based on past survey findings and the income potentialities of prospective homebuyers, it was estimated that 10 percent of the new one- and two-family home purchases and 12 percent of the existing home purchases would be on an all-cash basis. A final factor that would tend to reduce gross mortgage requirements, would be purchases of existing homes subject to existing mortgages, that is, where the buyer assumes the existing mortgage and a new one is not required. It was estimated, based on data from the Bureau of the Census' 1955-56 national housing inventory that 25 percent of the existing one- and two-family home purchases would be subject to existing mortgages.

Several factors making for increases have been incorporated in the estimates of gross mortgage credit requirements. These included rising loan-to-purchase price ratios, particularly in conventionally financed purchases; an allowance for second mortgages equal to about 4 percent of the amount of conventional first mortgages made in a year; and increases in average prices of homes purchased. The price increases of 3 percent per year for new houses and 1 percent per year for existing homes would reflect primarily increased quality and size of homes purchased. This is not in conflict with the assumption of stable prices which would apply to a house of a given size and quality. (In the case of new homes, the 3 percent annual price increase includes an allowance for rising land costs, which are not included in gross national product and would not be encompassed by the stable price assumption.)

The annual numbers of existing home purchases were estimated on the basis of past relationships between new and existing home purchases, as found in surveys of consumer finance that had been made for the Federal Reserve Board. Based on other available data, estimates were made of the relative proportions of home purchases and multifamily property construction and transfers that would be financed either with FHA-insured and VA-guaranteed, or with conventional mortgages. A moderate increase in the FHA-VA proportion was projected in the light of direct relationships in the past between high annual levels of home building and a relatively high proportion of home financing with FHA-insured and VA-guaranteed mortgages.

The aforementioned and other minor considerations entered into the derivation of estimates of annual gross residential mortgage credit requirements. The projected annual gross mortgage credit requirements ranged from about $28 billion in 1961 to about $49 billion in 1970. The latter are much greater than the net mortgage credit requirements (i.e., the gross requirements minus repayments on outstanding mortgages), the more significant measure to be considered. Available information on repayments of different types of mortgage loans (FHA-insured, VA-guaranteed, and conventional) held by different types of lenders was used to arrive at projected mortgage repayment ratios. Repayments include regular prepayments and partial or whole prepayments of principal, including complete repayments which occur upon transfer of ownership. Using the projected repayment ratios, the 1959 benchmark data on outstanding mortgage debt, and the estimated gross mortgage credit requirements-equal

to the annual amount of loans to be made-estimated annual net residential mortgage requirements were derived. They range from about $9 to $10 billion in the early years of the decade to about $14 billion in the last years of the decade, all in 1959 dollars.

The estimated annual net mortgage requirements, or net increases in residential mortgage debt projected for the sixties, do not rise to a level as high as the 1959 net increase in residential mortgage debt. The low level of projected annual net increases in mortgage debt during the early sixties is consistent with the assumed reduction from 1959 in the annual level of new starts. The projected net annual increases in mortgage debt or net mortgage requirements in the latter part of the sixties would not exceed previous peaks primarily because repayments from an increased total of outstanding mortgage debt would provide increasingly larger offsets to the gross mortgage credit requirements.

The total required net increase in outstanding residential mortgage debt during the decade ending with 1970 is estimated at $113 billion to bring the outstanding total up to $272 billion, in 1959 dollars. This is markedly lower than other estimates of net increases in mortgage debt which have been derived as estimated ratios of projected total residential construction expenditures, based on relationships between these two dollar series in the past decade.

In the fifties, however, there was a marked increase in mortgage interest rates, maturities, and loan-to-value ratios. These developments tended to work against the growth of mortgage principal repayments and the reduction of annual net mortgage fund requirements. It is doubtful whether mortgage interest rates will move to higher levels in the sixties. If they decline, mortgage principal repayments would tend to increase, since a larger proportion of average. debt service payments would go toward principal repayment. With regard to possible increases in home mortgage maturities it seems unlikely that they would increase nearly as much beyond the 25- and 30year terms in the sixties, as they increased to approach those terms in the fifties. Some further increases in loan-to-value ratios have been incorporated in the projected estimates of net mortgage requirements. All in all, therefore, a net increase in outstanding residential mortgage debt of between $100 billion and $125 billion (in 1959 dollars) during the 1961-70 decade would seem to be a reasonable projection. Nonhousing credit requirements

The extent to which funds may be available for mortgage financing, will depend in large measure upon the competitive demands for credit from the rest of the economy. As a prerequisite to judgments about the adequacy of mortgage credit, therefore, it is necessary to estimate the approximate net amounts of nonhousing credit that would be required by the nonfinancial sectors of the economy. The latter include the nonfinancial business sectors, corporate, noncorporate, and farm; governments; consumers and nonprofit organizations; and the rest of the world (i.e., U.S. net foreign investment).

This sector classification was adopted from the flow of funds accounts published quarterly by the Federal Reserve Board. From the same source, historical data and 1959 benchmark data were used to judge the distribution of sector credit requirements among different types of long-term and short-term credit. Some background and

benchmark data on various types of investment expenditures were taken from the national income accounts published by the Department of Commerce. Projected investment requirements for 1970 were generally taken from the projection of the American economy in 1970 prepared by the National Planning Association (in 1959 dollars). Investment requirements for intervening years were generally derived by interpolation between the 1959 benchmark data and the projected 1970 fiurges.

Projected estimates were developed of the annual net credit requirements that have to be met in the market. Excluded were the capital expenditures of financial institutions, generally financed out of their own capital and surplus funds, which will not affect the demand for funds in the market (although they will affect the supply). Also excluded from the net credit requirements are the funds borrowed by Government agencies which relend them to meet credit requirements that are already included in the analysis. Thus, net increases in funds borrowed by FNMA would be excluded because they would flow into mortgages for which the net requirements have already been calculated. Some of these exclusions and the absence of any net credit requirements by the Treasury, under the assumptions of a balanced Federal cash budget, may make the projected total net credit requirements seem low in relation to some estimates of credit expansion for

recent years.

In billions of 1959 dollars, the projected net credit requirements, by major categories, are as follows for 1961 and 1970:

TABLE 1A.-Projected net credit requirements, by major categories

The annual total amounts projected for intervening years range between those shown, rising moderately between 1961 and 1965, when net mortgage requirements are lower than in 1961, and more rapidly between 1965 and 1970 when the annual net mortgage credit requirements are greater. The projected annual net requirements for longterm nonhousing credit and for short-term credit increase each year. Net flows of cash savings and other loanable funds

Funds that will be available for the net credit requirements to be met in the market will come from the cash savings of consumers and nonprofit organizations, plus other funds invested in equity or credit instruments acquired in the market. These other investments are by corporate and noncorporate nonfinancial businesses, by fire and casualty insurance companies and by governments.

The concept of cash savings of consumers and nonprofit organizations that was adopted is the net acquisition of financial assets by that sector of the economy. It includes savings in institutions and

in private life insurance and pension funds, as well as credit and equity instruments, including mortgages, acquired in the market. It excludes savings in government funds that are used to offset credit requirements of governments. This concept of cash savings of consumers and other organizations is different from concepts under which an amount equal to the increase in consumer debt is deducted, or from concepts which include the value of the net increase of consumer investments in durable goods and other tangible assets. The types and amount of savings which would be included in all of these concepts probably is so large, however, that the stable relationship of savings to disposable personal income which has been found by authorities on savings, such as Raymond Goldsmith, would hold for all of them.

Therefore, based on the relationship shown by comparable data for the fifties, exclusive of the recession-affected years of 1954 and 1958, the net cash savings, or net acquisition of financial assets, was projected at 7.7 percent of projected disposable personal income, in constant 1959 dollars. (The projection of disposable personal income was derived on the basis of relationships during the fifties between gross national product, national income, personal income, and disposable personal income.) The total of such savings by consumers and nonprofit organizations, as projected, would rise from about $27 billion in 1961 to about $42 billion in 1970. If the savings rate is different, and there is likelihood that it would be, the results would differ from the model.

The projected annual flows of these savings were distributed among various types of institutional intermediaries who invest the savings funds and among direct consumer investment outlets. Distributive trends indicated by experience of the fifties were followed. Consequently, as projected, the largest proportion of net savings inflow would go into savings and loan associations. Private pension funds would be the second largest medium for savings. Private life insurance and commercial banks would be about equally important, and mutual savings banks would receive the smallest institutional share. Another source of savings available to meet net credit requirements in the market is the net increase in assets over the net increase in liabilities of financial institutions. They represent the undistributed profits or net surpluses of the financial institutions, and are reflected in the annual net financial investment by these institutions. In addition there are also the net financial investments of nonfinancial businesses, governments, and fire and casualty insurance companies. Annual net flows of funds from all of these sources were projected on the basis of trends indicated by data for recent years.

The projected totals of the net annual flow of funds into savings. intermediaries from consumers and nonprofit organizations plus the net acquisition of other financial assets by that sector and other components of the economy ranges from about $41 billion in 1961 to about $57 billion in 1970.

As a final step in building the hypothetical flow-of-funds model, the proportions of the projected net annual amounts of savings and other funds that would be made available to meet the mortgage and other credit requirements were estimated. It was assumed that the structure of interest rates would be about as it had been during the fifties, that the distribution of net savings inflows among different institutions