In other words, they represent the best approximations to actual building costs that are available. Bureau of Labor Satistics is the official source of Government statistics on residential construction, Since the adjusted BLS construction cost figures do not cover the cost of land nor the operative builder's overhead and profit, 33 percent has been added to the BLS construction costs to arrive at estimated sales prices. This means that 25 percent of the final sales price is estimated to be attributable to land, overhead, and profit. This is a generous allowance in view of the fact that profits and overhead are included in the BLS estimates of construction cost where houses are built on contract rather than by operative builders, and that profits and overhead are either small or nonexistent where owners themselves are the builders and no contractors or only a few special trade contractors are employed. A substantial volume of houses was built in 1947 on contract or by the owners themselves. Although no precise data are available, the Chief of the Construction Division of the Bureau of Labor Statistics, in his statement of January 12, 1948, before the Joint Committee on Housing, pointed out that "the owner-built house is becoming increasingly important in the suburban areas." Since only land needs to be added to construction costs in these cases to obtain an estimate of total cost to the owner, the average allowance of 33 percent means an allowance of about 40 percent for the remaining houses constructed by operative builders for sale, to cover their overhead and profit as well as land costs. The original BLS figures and the estimate of sales prices based thereon are shown in the following table: TABLE II.-Cost and price distribution of new single-family houses in the second quarter of 1947 Table II presents the Bureau of Labor Statistics data for the second quarter of 1947, this being the most typical period for housing starts in all sections of the country and the last one in which the Bureau made resurveys to obtain adjustments of the original costs reported to it. If the Bureau's data for the last quarter of 1946 and the first quarter of 1947 were used, they would show an even greater concentration in the lower cost and price brackets, and the comparison with family incomes would be more favorable. However, housing starts during the winter months are concentrating in the South where both construction costs and incomes are lower than in other parts of the country. The second-quarter figures therefore provide a fairer basis for comparison with incomes. COMPARISON OF INCOMES AND PRICES For the purpose of comparing the price of housing with the ability to pay for it, the widely accepted rule, that a family usually can afford to buy a home costing up to two and one-half times the annual income of the family, has been adopted. Thus, a family having an annual income of $3,000 is generally as sumed to be able to purchase and pay for a home costing $7,500. In view of the low financing charges for veterans and FHA home loans, with which about 50 percent of all new homes are financed, this formula may in fact be conservative. The next table and the accompanying chart bring together the data on family incomes and prices, which have been presented separately in Tables I and II. Since the available BLS figures do not permit a distribution of house prices on precisely a 2.5 ratio of price to income, the table gives the nearest possible distribution to this ratio. In all ranges, it will be noted, the range of price is below the maximum which families in the various income groups can be assumed to afford. The maximum price range is shown for comparison in the sixth column of the table. TABLE III.-Comparison of family income distribution and price distribution of new single-family dwellings Columns 1, 2, and 3 are from table II; columns 4 and 5 are from table I; column 6 is 21⁄2 times the family income brackets given in column 4. It thus becomes evident that, of the single-family houses built last year, 20 percent were in the price range of families with income below $2,000 who constituted 28 percent of the total families; 23 percent of the homes were in the price range of families with incomes between $2,000 and $3,000 who constituted 20 percent of the total; 25 percent were in the price range of families with incomes between $3,000 and $4,000 who constituted 18 percent of the total; 18 percent were in the price range of families with incomes between $4,000 and $5,000 who constituted 13 percent of the total; and only 14 percent were in the price range of families with incomes of $5,000 or more who constituted 21 percent of the total. If, as is probable, only a few of the new houses in the lowest price class were purchased by families with annual incomes below $1,000, then the 21 percent of the houses in this class could be considered to have been built for the 15 percent of all families-those having between $1,000 and $2,000 annual income who formed the lowest income group actively in the market. Excepting the relatively small group of families having incomes of less than $1,000 a year, which cannot be expected to be in the market for newly constructed housing, the one group for which new houses were not provided in proportion was that with income of $5,000 a year and over. The highest income group, instead of getting all, or even a disproportionately large share of the new housing, as frequently is charged, actually was provided with less than its reasonable allotment. OTHER SUPPORTING DATA The above findings are in a broad way confirmed by evidence from other Government statistics. According to the Economic Report of the President of January 14, 1948, the average income of all American families in 1946 was $3,806. This, it should be noted again, includes the incomes of farm families and of individuals living by themselves. The average income of normal nonfarm families of two or more persons must have been considerably higher. Again, on the rule that a family normally can afford to buy a house costing up to 21⁄2 times its annual income, the average American family was able to buy a house priced at $9,515. Acording to a report of the Veterans' Administration, the average selling price of new houses bought by veterans with the aid of GI home loans was $8,200,5 or $1,315 less than the amount the average family could afford. New houses bought by veterans represent a fair sample of all new houses built during 1947. The Veterans' Administration states that nearly half of the home loans that it guarantees or insures are for newly built houses, and since these loans totaled about 500,000 last year, 250,000 new houses must have been involved. Table 2, p. 18. The average News release of the Veterans' Administration, dated November 30, 1947. is based on data for the 4-month period from May through August. The same release stated that nearly one-half of all VA-guaranteed loans were on new homes. In other words, they represent the best approximations to actual building costs that are available. Bureau of Labor Satistics is the official source of Government statistics on residential construction. Since the adjusted BLS construction cost figures do not cover the cost of land nor the operative builder's overhead and profit, 33 percent has been added to the BLS construction costs to arrive at estimated sales prices. This means that 25 percent of the final sales price is estimated to be attributable to land, overhead, and profit. This is a generous allowance in view of the fact that profits and overhead are included in the BLS estimates of construction cost where houses are built on contract rather than by operative builders, and that profits and overhead are either small or nonexistent where owners themselves are the builders and no contractors or only a few special trade contractors are employed. A substantial volume of houses was built in 1947 on contract or by the owners themselves. Although no precise data are available, the Chief of the Construction Division of the Bureau of Labor Statistics, in his statement of January 12, 1948, before the Joint Committee on Housing, pointed out that "the owner-built house is becoming increasingly important in the suburban areas." Since only land needs to be added to construction costs in these cases to obtain an estimate of total cost to the owner, the average allowance of 33 percent means an allowance of about 40 percent for the remaining houses constructed by operative builders for sale, to cover their overhead and profit as well as land costs. The original BLS figures and the estimate of sales prices based thereon are shown in the following table: TABLE II.-Cost and price distribution of new single-family houses in the second quarter of 1947 Table II presents the Bureau of Labor Statistics data for the second quarter of 1947, this being the most typical period for housing starts in all sections of the country and the last one in which the Bureau made resurveys to obtain adjustments of the original costs reported to it. If the Bureau's data for the last quarter of 1946 and the first quarter of 1947 were used, they would show an even greater concentration in the lower cost and price brackets, and the comparison with family incomes would be more favorable. However, housing starts during the winter months are concentrating in the South where both construction costs and incomes are lower than in other parts of the country. The second-quarter figures therefore provide a fairer basis for comparison with incomes. COMPARISON OF INCOMES AND PRICES For the purpose of comparing the price of housing with the ability to pay for it, the widely accepted rule, that a family usually can afford to buy a home costing up to two and one-half times the annual income of the family, has been adopted. Thus, a family having an annual income of $3,000 is generally as sumed to be able to purchase and pay for a home costing $7,500. In view of the low financing charges for veterans and FHA home loans, with which about 50 percent of all new homes are financed, this formula may in fact be conservative. The next table and the accompanying chart bring together the data on family incomes and prices, which have been presented separately in Tables I and II. Since the available BLS figures do not permit a distribution of house prices on precisely a 2.5 ratio of price to income, the table gives the nearest possible distribution to this ratio. In all ranges, it will be noted, the range of price is below the maximum which families in the various income groups can be assumed to afford. The maximum price range is shown for comparison in the sixth column of the table. TABLE III.—Comparison of family income distribution and price distribution of new single-family dwellings Columns 1, 2, and 3 are from table II; columns 4 and 5 are from table I; column 6 is 21⁄2 times the family income brackets given in column 4. It thus becomes evident that, of the single-family houses built last year, 20 percent were in the price range of families with income below $2,000 who constituted 28 percent of the total families; 23 percent of the homes were in the price range of families with incomes between $2,000 and $3,000 who constituted 20 percent of the total; 25 percent were in the price range of families with incomes between $3,000 and $4,000 who constituted 18 percent of the total; 18 percent were in the price range of families with incomes between $4,000 and $5,000 who constituted 13 percent of the total; and only 14 percent were in the price range of families with incomes of $5,000 or more who constituted 21 percent of the total. If, as is probable, only a few of the new houses in the lowest price class were purchased by families with annual incomes below $1,000, then the 21 percent of the houses in this class could be considered to have been built for the 15 percent of all families-those having between $1,000 and $2,000 annual income who formed the lowest income group actively in the market. Excepting the relatively small group of families having incomes of less than $1,000 a year, which cannot be expected to be in the market for newly constructed housing, the one group for which new houses were not provided in proportion was that with income of $5,000 a year and over. The highest income group, instead of getting all, or even a disproportionately large share of the new housing, as frequently is charged, actually was provided with less than its reasonable allotment. OTHER SUPPORTING DATA The above findings are in a broad way confirmed by evidence from other Government statistics. According to the Economic Report of the President of January 14, 1948, the average income of all American families in 1946 was $3,806.' This, it should be noted again, includes the incomes of farm families and of individuals living by themselves. The average income of normal nonfarm families of two or more persons must have been considerably higher. Again, on the rule that a family normally can afford to buy a house costing up to 21⁄2 times its annual income, the average American family was able to buy a house priced at $9,515. 5 Acording to a report of the Veterans' Administration, the average selling price of new houses bought by veterans with the aid of GI home loans was $8,200, or $1,315 less than the amount the average family could afford. New houses bought by veterans represent a fair sample of all new houses built during 1947. The Veterans' Administration states that nearly half of the home loans that it guarantees or insures are for newly built houses, and since these loans totaled about 500,000 last year, 250,000 new houses must have been involved. Table 2, p. 18. The average 5 News release of the Veterans' Administration, dated November 30, 1947. is based on data for the 4-month period from May through August. The same release stated that nearly one-half of all VA-guaranteed loans were on new homes. According to the monthly reports of the Federal Housing Administration, the average mortgage amount per family unit insured during 1947 under section 603 of the National Housing Act was $6,745. Under this section of the act, the FHA insures loans on new one- to four-family dwellings up to 90 percent of the FHA estimate of current necessary cost. Allowing for cases in which the loan percentage was less than the statutory maximum of 90 percent and also allowing for differences between the FHA cost estimate and the actual sales price, the average price of new FHA-financed dwelling units probably was in the neighborhood of $8,000. On this basis, the average FHA mortgage would be 84.3 percent of the average sales price. An average sales price of $8,000 would be $1,515 less than the amount the average American family could afford to pay for a new house. Here again, new houses financed with FHA loans represent a fair sample of all new houses built during 1947. Almost 180,000 new units in one- to four-family dwellings were placed under construction last year under the FHA program, or more than 20 percent of all new units of this type. The FHA generally operates neither in the luxury market nor in the lowest segment of the market. These corroborating data from the Veterans' Administration and the Federal Housing Administration are completely independent from the figures of the Bureau of Labor Statistics, which were used in the previous analysis. They strongly support the other evidence of the broad market served by private home building. EFFECT OF GEOGRAPHIC DISTRIBUTION A large proportion of new housing in 1947 was built in the so-called rural-nonfarm areas, outside of the political boundaries of cities, where land and construction costs are generally lower than in cities. The question might be raised, therefore, that the findings of this survey could be explained by an undue concentration of new home building in rural sections far away from the places of employment for the majority of city dwellers. This, however, is not the case. About two-thirds of all new nonfarm dwelling units started in the first three quarters of 1947 were located within metropolitan areas containing cities of 50,000 population and over, or in commuting distance for the vast majority of employees and workers. This proportion is approximately the same as before the war. Since the data used in this study are national data, they do not necessarily apply to any particular areas or cities. However, the available figures even for large industrial areas known to be high-cost areas refute the theory that new private housing is serving only the upper-income families, as the following table shows: TABLE IV.-Cost and price distribution of new single-family houses in industrial areas, second quarter of 1947 Even in the large industrial areas, 57 percent of the houses were in the price ranges below $9,642, which corresponds closely to the figure of $9,515 previously calculated as the average price that the family with average income could afford. Yet, it is generally recognized that family incomes in the large industrial centers are higher than the national figures used in this comparison. The sample of industrial areas from which the Bureau of Labor Statistics' figures on construction costs are drawn comprises 27 centers including New York, Statement of the Chief of the Construction Division of the Bureau of Labor Statistics before the Joint Committee on Housing, table 5. |