The purpose of this appendix is to present in one convenient place the data, sources, and description of the methods used in deriving the data supporting certain tables and charts presented in the body of the report. In some cases the materials given in the text required extensive discussion and supporting statistics, and in such cases a separate appendix was written to cover them. In other cases a single reference was sufficient to indicate the source. The cases requiring more than a reference but not justifying a separate appendix are brought together in this appendix. The order of the materials corresponds to the order in which they are discussed in the text.

1. Estimate of the Amount of Food Stored in New York City

In chapter I an estimate is given of the food supply in the metropolitan area of New York. A conditional estimate of the amount of food stored in New York City may be made for the end of the year 1929 from data published by the Bureau of the Census. The estimate was made as follows: Total sales by retail stores "in the food group" and for the classification "restaurants, cafeterias, and eating places" were added to sales by wholesalers in the same classification to ultimate consumers at regular retail prices. The figures are shown in table I.

TABLE I.-Retail sales of food in New York City for the year 1929 [Millions of dollars]

Source: Fifteenth Census of the United States, wholesale figures from Vol. II, Wholesale Distribution, p. 1007, et. seq.; retail figures from Vol. I, Retail Distribution, part 3, 3. 215, et. seq.

From the same source, estimates of value of stocks on hand were made, value of wholesalers' stocks were marked up 10 percent and added to value of retailers' stocks; this total was then marked up 20 percent in order to get an estimate of the value of stocks in terms of consumers' prices. Computations are shown in table II.

1 Appendix 18 was prepared by Louis J. Paradiso, assisted by Grace W. Knott, Paul A. Fischer, Ezra Glaser and James Arnold; Marion Tolles also assisted in the preliminary research in connection with section 5 on wealth.

228. 4

Total at cost to retailer.

Total stocks_

Plus 20-percent mark-up

Total food stocks at prices to consumers_

1 These percentages are rough estimates covering sales of all foods assuming all sales went through food stores rather than restaurants. They are in agreement with the preliminary estimates obtained by the Department of Commerce investigations into average wholesale and retail mark-up in New York City.

Source: Fifteenth Census of the United States, wholesale figures from Vol. II, Whole sale Distribution, p. 1007, et. seq.; retail figures from Vol. I, Retail Distribution, part 3, p. 215, et. seq.

Total sales for the year 1929 were 1,426.9 millions of dollars, and the value of food stocks at prices to consumers held at the end of the year were, roughly, 228.4 millions of dollars. Thus, 16 percent of the year's sales were held in stocks at the date the survey was made. This is equivalent to about 58 days' or approximately 2 months' supply of food on hand.

Approximately the same result was obtained by making use of the findings of a report by the Port of New York Authority called Food Supply of the Port of New York District for the metropolitan district in 1922. The length of time of the supply for the various items of food given in this report varied from meat, of which there was only a week's supply in storage, to eggs, of which there was at most a 3 months' supply. On the average, it was found by a study of the detailed items that a 45 days' supply of food was on hand in the New York metropolitan district in 1922.

The estimate of a 2 months' supply is a conservative one in view of the fact that there is such a great variation in the supply of specific food items. Milk is shipped in daily and fresh eggs and meat at short intervals. Canned goods and packaged fancy groceries on the other hand might be stored in wholesalers' warehouses for long periods. Furthermore, the supply of some specific commodities varies greatly with the season

of the year, so that the estimate of the food supply on hand would be different for the different months of the year.

2. Loss in Potential Real National Income Due to Depression Unemployment of Men and Machines, 1930-37

In chart I of chapter I, real national income produced in the United States is shown for the years of the period 1920 to 1937. The line showing potential real national income was obtained by connecting with a compound interest curve the average real national income produced for the years of the period 1923-29 (this average being centered at 1926) with an estimated real national income that could have been expected in 1938 at practical full employment. For the purposes of this chart, practical full employment in 1938 was assumed to correspond to a residual unemployment of 2 millions, using an estimated labor force in 1938 of 54.5 millions.2 The estimated potential real national income in 1938 corresponding to practical full employment is 103.2 billions of 1929 dollars. This figure was derived from total employment calculated at different assumed levels of consumer income shown in table II of the report Patterns of Resource Use, National Resources Committee. The index of industrial production corresponding to a residual unemployment of 2 millions was interpolated from table II of the Pattern report to be approximately 147. The real national income produced corresponding to the index of industrial production of 147 was calculated from the following relationship:

This relationship was derived by the method of least squares, using the logarithmic form, from the data on real national income given in the table below and the Federal Reserve Board index of industrial production for the years of the period 1921 to 1937. The average percent residual 3 for the years of the entire period is 1.4 percent, thus indicating a fairly close relationship between real national income produced and industrial production.

The curve representing potential real national income, obtained by connecting the 1923-29 potential real national income with the calculated 1938 potential real national income, is given by the following compound interest formula:

National income produced=73.1 (1.029) year-1926
(Billions of 1929 dollars)

* This estimate is based on a projection of unpublished data on labor force estimated by the National Research Project of the Works Progress Administration. For the data on labor force, see Patterns of Resource Use, National Resources Committee, 1938, appendix 2, table I.

3 The difference between real national income as calculated from the formula and the actual expressed as a percent of the calculated values.

1 From 1919-28 based on national income as given in National Income and Capital Formation, National Bureau of Economic Research, p. 8. This was made comparable to the Department of Commerce estimates (given in Survey of Current Business, June 1938) of national income produced by adjusting for net imputed rent, net government savings, depreciation, and changes in inventory.

2 National Bureau of Economic Research, Bulletin 59: figures for later years furnished by the Bureau: based on prices of capital goods weighted by 1 and Bureau of Labor Statisttics cost of living index weighted by 9.

3 Based on the following compound interest curve obtained from the average of real national income produced for the years of the period 1923-29 (centered at 1926) and the estimate of potential national income in 1938 of 103.2 billions of 1929 dollars associated with a residual unemployment of 2 millions-see text for basis of calculation.

Potential real national income less actual real national income produced. 5 Due to depression unemployment of men and machines, 1930-37.

3. Major Items of Social Expenditures, 1935

The major items of social expenditures shown in chapter II, table IV, are as follows:

The figures for Federal Government expenditures were derived from the 1938 Budget of the United States (data for 1935–36). These expenditures were allocated to the functional break-down given below, excluding the items which did not fall into the categories used.

The figures for State governments were derived by the use of estimates in the report of the Twentieth Century Fund, Studies in Current Tax Problems, 1937, showing expenditures for certain functions for 1935-36. A large item "other" was allocated among the remaining functions on the basis of the ratios existing in 1932 as shown in the Census report, Financial Statistics of States and Local Governments, 1932.

The figures for local governments were derived by applying to the 1932 data for all local governments the percent changes from 1932 to 1936 shown in various expenditures of 39 large cities. The trend for these 39 cities showed the same relationship as shown for 94 cities between 1932 and 1935. Reports on the 94

cities cover approximately 60 percent of all local governments. The figures on local government expenditures are probably too low.

The following are the categories and sources used:

1. Education-includes expenditures for all schools, including books and supplies; and the promotion of education. The figure for private agencies' contribution to education represents the income received from endowments to institutions of higher learning as reported by the U. S. Office of Education for the school year 1935-36. Private and parochial schools are not included. Individual expenditures are the total expenditures of families and single individuals for education as shown in the National Resources Committee report, Consumer Expenditures in the United States.

2. Health and sanitation-includes expenditures for prevention and treatment of diseases; medical and dental work for school children; food regulation; sanitary inspection; and disposal plants. Individual expenditures are the total expenditures of families and single individuals for medical care as shown in the National Resources Committee report, Consumer Expenditures in the United States.

3. Recreation-includes expenditures for all recreational facilities, including parks, playgrounds, museums, including projects of the Works Progress Administration in these fields. Individual expenditures are the total expenditures of families and single individuals for recreation as shown in the National Resources Committee report, Consumer Expenditures in the United States.

4. Reading includes expenditures for all libraries for use of the public; also payments to private library associations. Private agencies' contribution was derived as follows: The latest figure for total amount of endowments from individuals to libraries is for 1929. The Bureau of Library Service of the United States Office of Education furnished the figure of $95,820,560 as the amount of permanent endowment funds. A weighted yield on Treasury, municipal and corporate bonds in 1935 was calculated and applied to the amount of permanent endowments. Individual expenditures represent the amount of total expenditures made by families and single individuals for reading as shown by the National Resources Committee report, Consumer Expenditures in the United States.

5. Highways--includes all expenditures for operation and maintenance of public roads, bridges, etc., as reported by the Bureau of Public Roads, United States Department of Agriculture.

6. Charities, hospitals and corrections-includes administrative expenses for supervision of relief and of institutions; expenditures for general hospitals maintained by Federal, State, and local Governments (excluding those for care of insane and feeble-minded) and

probation boards. This does not include payments to persons on relief or living expenses of persons in institutions. Private agencies' contribution was derived as follows: The Final Report of the Committee on the Costs of Medical Care, 1932, showed the source of funds for hospitals in 1929 which included an item of 54 million dollars derived from philanthropy. To this figure was added an estimated sum of 65 million dollars as contributions made through private relief agencies. According to a report of the Works Progress Administration, Trends in Relief Expenditures, 1910-1935, the total amount of public relief in 1935 amounted to approximately 2,300 million dollars. The ratio of private relief to public relief in 120 urban areas was 2 percent and in 385 rural towns, 4 percent. Weighting these ratios according to total population in urban and rural areas, a ratio of 2.9 percent was derived and applied to the total public expenditures. This figure for private agencies includes payments made to clients.

7. Interest payments-includes expenditures for interest payments on outstanding debt.

8. Churches-This figure represents the amount of gifts made by families and single individuals to churches in 1935-36, based on data from the National Resources Committee report, Consumer Expenditures in the United States. Part of the expenditures made by churches from these funds go to foreign missions. In addition to these funds churches derive income from endowments and real estate holdings. No data were available to permit an estimate of these amounts. It has been assumed that these two figures roughly offset one another. Part of contributions to churches go to support of schools and could properly be included under education.

4. Amount and Proportion of Goods and Services by Degree

of Durability, 1919-35

In chart IX of chapter II and in subsequent chapters the values of goods and services are presented by years according to the degree of durability. The data on the value of goods were obtained directly from Kuznets, Commodity Flow and Capital Formation. The value of services was derived by the National Resources Committee. In this section a description of the method used for deriving the service series is given and a table is presented of the amount and proportion of goods and services by degree of durability.

The series on value of services was derived by estimating the value of services in current dollars and deflating the resulting series by a price index for services. The value of services in current dollars was computed from seven component series and the price index for services was calculated from a weighted average of seven price indexes.

The following series were used as components of the services series:

The first series consists of income produced by the segments included in the services to the consumer. These include private education, personal service, professional service, recreation and amusement, domestic service, and miscellaneous consumer services. The data for income produced for the years of the period 1929-37 were obtained from the United States Department of Commerce, Income Section. For the years of the period 1919-29, income originating from services was obtained from Kuznets, National Income and Capital Formation, 1919-35, page 67. These data were adjusted to the Department of Commerce series by the ratio of the two series in 1929.

The second series consists of income from residential telephone service. The data for the years of the period 1926-36 were derived from the Federal Communications Commission, Third Annual Report, and represent all types of operating income. For the period 1935–36, 61.3 percent of total operating revenue was from residential telephones. This percentage was assumed to hold for earlier years. The data for the years of the period 1919-25 were obtained by linear extrapolation along a trend line.

The third series is income from residential electric power service. This series was calculated by multiplying the rate per kilowatt-hour by total kilowatt-hours utilized by residential consumers. Rates Rates per kilowatthour were computed from the Federal Power Commission, Trends in Residential Rates, 1924-36. Rates for 25, 100, and 250 kilowatt-hours were given weights of 7, 2, and 1, respectively, these weights being based on the relative importance of the quantities consumed. The data for the years 1919-23 were obtained by linear extrapolation along a trend line. The data for the years of the period 1926-36 on total kilowatt-hours consumed were obtained from the Edison Electric Institute, Statistical Bulletin, No. 4, New York, 1937. For the other years the data were based on Moody's Public Utilities and quoted from the Electrical World.

The fourth series is railway passenger revenue. Data were obtained from Interstate Commerce Commission, Statistics of Railways in the United States, 1936. Data for the year 1936 were obtained from Interstate Commerce Commission monthly statement, Revenue Traffic Statistics of Class I Railways, December

1936.

The fifth item is motor-bus revenue. The data for the years 1927-36 were derived from Bus Transportation; the data for the years 1919-26 were obtained by linear extrapolation along a trend line.

The sixth series is electric-railways revenue. This was computed from average fare (described below) multiplied by the number of passengers. The source

for number of passengers for the years 1917, 1922, 1927, and 1932 is the census of Street Railways and for the intermediate years the source is the American Transit Association. These data are reported in Moody's Public Utilities.

The last series is revenue from private first-class mail. These data were obtained from the Statistical Abstract of the United States and based on the annual reports of the Postmaster General.

The above seven series were summated and the resulting series was deflated by a price index for services which was constructed from the following component series: (1) Index of price of services to the consumer, (2) residential telephone rates, (3) residential electric power rates, (4) railroad passenger rates, (5) motor bus rates, (6) electric railway rates, and (7) first-class mail rates. These various indexes of price were weighted by the following weights: 9722, 714, 600, 876, 350, 1149, and 627, respectively. These weights were based on the relative values of the various service items in 1929.

The following is the description of sources and methods used for getting the price series:

The index of price of services to the consumer is represented by the Bureau of Labor Statistics' cost of living for "miscellaneous," as reported in Standard Statistics, for the years of the period 1920-36. This series includes the cost of such items as reading materials, tobacco, organization dues, medical care, car fare, drugs, toilet articles, etc., but it excludes food, clothing, housing, fuel and lighting, and house furnishings. For the year 1919, the index was estimated from the percent change shown in the two years 1919 and 1920 by the National Industrial Conference Board's index of sundries, as reported in Standard Statistics, which is a component of the cost of living index computed by that organization.

Residential telephone rates were computed from data compiled by the Wisconsin Public Service Commission for the years of the period 1926-36. For the years of the period 1919-25 the series was obtained by linear extrapolation along a trend line.

Residential electric power rates, railroad passenger rates, and motorbus rates were obtained from the same sources as the revenue series for the corresponding items. Residential electric power rates and motorbus rates were extrapolated along trend lines for those years for which data were not available.

Electric railway rates consist of an unweighted average of monthly cash fares for street railways or bus service in cities of 25,000 or more population. The number of cities was not constant-320 cities in 1932 and 268 cities in 1936. These data were compiled by the American Transit Association, formerly the American Street Railway Association, and published in the

4 See appendix IV, table V, of this report.