C\3>, intA data are for calendar years, diary estimates are for fiscal years, July 1972-June 1374. 2 Preliminary. 3 No adjustment for footwear. 22 Table 5-2. -Aggregate Expenditures for Clothing and Footwear: Quarterly Panel and Diary Operation Compared to Independent Data Sources, 1972 and 1973 [Billions of dollars] Source 1972 Clothing Footwear 1973 Clothing Footwear Survey estimates Quarterly panel Diary operation Personal Consumption Expenditures (PCE) (Gross National Product accounts) .... Census of Business Merchandise line data . $30.9 31.6 45.2 40.9 $5.7 6.7 8.9 7.7 $34.3 34.1 50.3 (NA) $6.2 6.9 10.0 (NA) (NA) = Not available. Survey estimates may or may not include sales taxes which are explicitly included in the independent estimates. Thus, the actual differences might be a few percentage points smaller than indicated. Note. — Coefficients of variation of quarterly panel estimates for a single year are about 1 percent both for clothing and footwear. For the diary oper- ation, the comparable coefficients are about 3 percent for clothing and 4 percent for footwear. Table 5-3.— Aggregate Estimates of Clothing Expenditures, by Major Category: Quarterly Panel Compared to Independent Data Sources, 1972 and 1973 [Millions of dollars] Category Quarterly panel estimates 1972 1973 Ratio of quarterly panel estimates to: Personal Consumption Expenditures (GNP accounts) 1972 1 Adjusted Nat'l Consumer Panel Data (Market Research Corporation of America) 1973 2 Coats, suits, and tailored jackets Untailored jackets and sweaters Dresses and skirts Trousers and Slacks Shirts and blouses Underwear Nightwear Hosiery Hats, gloves, and accessories . . $7,427.1 1,824.4 4,203.5 5,319.8 3,787.1 1,937.4 1,151.2 2,020.0 1,276.4 $8,547.8 2,335.2 4,030.7 6,096.6 4,438.1 1,961.8 1,287.6 1,954.5 1,426.1 1.08 .63 .71 .84 .68 .59 .71 .88 .52 1.28 .71 .76 .95 .78 .61 .83 .93 (NA) (NA) = Not available. 'Projected by author from 1967 PCE estimates, using Census of Manufactures and census foreign trade statistics for 1967 and 1972 (see appendix A). 2 Computed from basic panel data using adjustment factors supplied by MRCA (see appendix A). Note. — All proportions in last two columns differ significantly from unity. Table 5-4.-Aggregate Expenditures for Clothing and Footwear: Quarterly Panel Estimates Based on Data for All Months and on Reported Purchases in Final Month of Each 3-Month Period, Compared to Independent Data Sources, 1972-1973. [Millions of dollars] Category Quarterly panel estimates Based on data for all months Based on reported purchases in final month of each 3-month period Ratio of quarterly panel estimates to independent estimates' for: Data for all months from panel "Final month' data from panel Clothing— total 2 Coats, suits, and tailored jackets Untailored jackets and sweaters Dresses and skirts Trousers and slacks Shirts and blouses Underwear Nightwear Hosiery Hats, gloves, and accessories . . Footwear— total $32,575.3 7,987.7 2,080.0 4,117.1 5,708.2 4,112.6 1,949.6 1,219.4 1,987.3 1,351.3 5,959.9 $38,613.9 9,138.5 2,353.9 4,987.7 7,044.5 5,121.6 2,338.0 1,473.6 2,325.4 1,638.2 7,313.7 .68 1.20 .67 .73 .90 .73 .60 .77 .85 3.53 .63 .81 1.37 .76 .89 1.11 .91 .72 .93 1.08 3 .64 .77 'For total clothing and footwear, independent estimates used for comparison are average of Personal Consumption Expenditures (PCE) estimates for 1972 and 1973. For clothing categories, Independent estimates are average of PCE estimates for 1972 and National Consumer Panel (MRCA) data for 1973 which were used in Table 5-3. 2 Detail will not add to total because certain categories are omitted for which comparisons could not be made. 3 Independent estimate used for both years was PCE estimates for 1972. CHAPTER 6 Expenditures for Major and Minor Appliances and Equipment This chapter deals with a wide range of household appliances and equipment, including both large and costly products, such as refrigerators, washing machines, and television sets, and small, relatively inexpensive products, such as toasters, radios, and fans. Although altogether they account for only a small pro- portion of total consumer expenditures— some $15 to $20 billion at the time of the survey— purchases of these items appear to be particularly affected by cyclical fluctuations in economic activity and are, therefore, valuable indicators in studying economic trends. Because particular appliances are purchased infrequently by in- dividual families, sampling variances are relatively large com- pared to those for other kinds of expenditures. For this reason, the quarterly panel— in which a full year's experience is obtained for each family— was regarded as more realistic for this infor- mation than the diary operation, in which only two weeks' coverage is obtained for any one respondent. A special pro- cedure known as the "inventory" approach, which was dis- cussed in Chapter 1, was used for this purpose in the inter- view panel. Instead of inquiring directly about expenditures for a given period, respondents were asked at the first interview about possession of the articles in question. If any such items were present, the date of acquisition was determined and, if within the previous year, the cost and a variety of other characteristics were recorded. This inventory was updated at subsequent visits with inquires about any new acqui- sitions either for the respondents' own use or as gifts for per- sons outside the family. The updating was done only once, at the end of the annual cycle, for major appliances. It was carried out twice, at semiannual intervals, for small appliances. No special procedures were followed for this class of expen- ditures in the diary operation. Because of the higher priority accorded other categories, appliances were not specifically mentioned on the diary form but were relegated to various catchall sections. Comparisons With Independent Estimates Previous efforts to measure purchases of and expenditures for appliances and other durable commodities of that type have met with mixed success. Because of the importance and cost of many of these products, respondents tend to remember pur- chases fairly completely for reasonable recall periods. How- ever, there is also a considerable tendency for respondents to "telescope" expenditures that actually occurred in a prior period. This problem was controlled in the present survey by using the "bounding" procedure. One of the problems in appraising the survey results is that much of the available independent data, whether from Govern- ment surveys or market research, are based on open-ended inter- views and are subject to telescoping of this kind. It was decided, therefore, to depend principally in evaluating the survey results on estimates from the Personal Consumption Expenditures (PCE) derived from the Gross National Product accounts. How- ever, in order to obtain estimates in sufficient detail for the survey date, it was necessary for the author to update the PCE data from 1967 to 1972, using the method described in ap- pendix A. In addition, an adjustment had to be made in each category to reflect appliances included as part of home pur- chases. The resultant estimates, therefore, can be considered only very rough approximations of level. A comparison of the survey data from the quarterly panel with the corresponding PCE estimates is presented in table 6-1 for new appliance expenditures. Two different methods were used to develop estimates from the survey results. To under- stand these methods, it will be necessary to elaborate a bit further on the survey procedure. When the initial "inventory" was obtained, items on hand were differentiated into two cate- gories—those which had been purchased by the family for its own use (including those included as part of a home purchase) and those the family had received as gifts from persons outside the household. As the inventory was updated during the course of the survey year (at semiannual intervals for small appliances, annually for the major ones), newly acquired items were differ- entiated into the same two classes. In addition, questions were asked each time about appliances the family may have pur- chased as gifts to be given to persons outside the household. This was done because it was believed that the most accurate expenditure estimates could be derived from purchases actually made by the family— whether for their own use or as gifts for others— because the family could more accurately report the prices paid. This combination of procedures is the basis for the first set of estimates (method 1 ) in table 6-1 . It is evident, however, that an expenditure estimate can also be made by combining purchases made by the family for its own use and the value of gifts they received from others. This com- bination is shown as the second set of estimates (method 2) in table 6-1. The problem is that the value of gifts received 23 24 cannot always be reported by respondents since price tags are obviously removed in advance in most cases. In fact, it was necessary to impute the value at the data-processing stage for about 15-20 percent of items reportedly received as gifts, whereas this was a negligible problem for those actdally pur- chased by the family. Of course, since gifts received from others and gifts given to others are merely different measures of the same thing, the two estimates should theoretically be identical. The fact that they differ in some instances reflects a rather interesting reporting phenomenon. For nearly every appliance class, the estimated value of gifts received from persons outside the household — which was derived from the inventory of items on hand at the time of interview— exceeded the estimated expenditures reported by the family for gifts given to persons outside the household, which was based on their recall of purchases made during the reference period. This appears to substantiate the notion that the inventory approach used in the survey pro- vides a more complete reporting of acquisitions than the cus- tomary technique of asking persons retrospectively about their expenditures for a given period. 1 In any event, the two types of estimates do not differ materially for major appliances, which are seldom received as gifts. More importantly, allowing for sampling errors and the considerable uncertainties about the independent estimates, both methods provided results that are reasonably consistent with the cor- responding PCE figures. Except possibly for television sets, there is no clear directional bias, either positive or negative, for the group of major appliances as a whole. Moreover, in the case of television, there is a good possibility that the PCE estimates may be exaggerated. 2 Other evidence, such as esti- mates from the Census Bureau's Annual Housing Survey, indi- cates levels much closer to those of the quarterly panel data. More noticeable differences between the two methods were observed for certain of the minor-appliance categories. The difference is especially large for small kitchen appliances, about half of which were acquired as the result of gifts. In this case, the first method, encompassing gifts given to others, fell mark- edly short of the PCE level. The estimate for method 2, which included gifts received from others, was still perhaps a bit low, but it was much closer to the PCE figure. For this product class, the aggregate value of gifts received (not shown separately in the table) was about $350 million compared to only $150 million for gifts reported given to others. For photographic equipment, the other product class showing a large disparity relative to the PCE level, the use of the second 1 The differences between the two methods of estimating gifts was reflected in the number of units involved as well as in aggregate ex- penditures, so that differences in reporting the value of individual gifts was not a prime factor. 2 Special difficulties were encountered in developing the 1972 PCE estimates for television sets because of major changes in the mix of products within the category and the sharply increased importance of imports. Similar problems emerged in the case of sewing machines, al- though, in that case, these problems apparently resulted in an under- statement in the PCE estimate. method does not appreciably close the gap. A likely explanation of the deficiency in this instance rests with the multiple posses- sion of cameras and related items by various family members. There was probably insufficient attention in compiling the "inventory" to the likelihood of multiple ownership of this kind, and the survey respondent (usually the homemaker) probably overlooked some of the equipment owned by individ- ual members. Comparison With Diary Results Although there was little intention of using diary results to measure appliance expenditures, it may be useful to examine the data to see whether or not they confirm previous opinion or experience in this field. Aside from problems of sampling variance, recordkeeping for short periods has been regarded as deficient in other ways for larger expenditures of this kind. One problem, especially for major durables, is determination of when an item is actually acquired— whether at the time a deposit is made, a contract signed, delivery received, payment com- pleted, etc. With a short diary period, uncertainties of this kind could result in postponing entries beyond the period of coverage. Estimates of aggregate annual expenditures from the diaries as compared to the quarterly panel are presented in table 6-2. Although sampling variances preclude definitive judgments, there appears to be a fairly consistent pattern of lower levels for the diary operation. Although some improvement might be possible by according greater attention to these items, the previous concern about likely understatement in diary-based estimates in this sector, while not completely confirmed, is certainly not allayed by these results. Use of Prior Year's Data With the large samples used in the 1972-73 survey, the variances for most of the major appliances were relatively modest (coef- ficients of variation of 4 to 8 percent for a single year). Even for individual small appliances, the coefficients were not excessive for a one-time survey. However, for a continuing survey pro- gram with smaller samples and the need for measuring short- term changes, there is a question as to whether the statistical reliability will be sufficient for the various purposes entailed. One possible means of improving the reliability, almost without cost, would be to take advantage of the data obtained for prior years in the course of the inquiry. Under the inventory ap- proach described earlier, all items currently on hand are record- ed along with their date of acquisition. For those acquired in the year preceding the starting date for the survey (e.g., 1972 in the case of the 1973 quarterly panel), descriptive and cost data comparable to that for current acquisitions were also re- corded. As a consequence (in addition to the different ways of estimating gifts), two separate estimates were available for 1972, one based on current acquisitions as determined in the 1972 quarterly panel operation and one based on acquisitions in the previous year as reported in the 1973 quarterly panel. In a con- tinuing survey operation, two such estimates would be available for each year. Since there is very little, if any, correlation in the 25 purchase of these items by the same families in successive years, the two estimates could be combined with at least the same statistical reliability as the data from two independent samples of the same size. 3 The effective sample size for these expenditures would, thereby, be doubled and the sampling variances reduced commensurately. One problem with such a procedure is that there would be some gap in the timing with which the two separate estimates become available in a continuing program. This gap could possibly be accommodated by the common practice of issuing a prelimi- nary estimate at the outset and a final estimate when the second set becomes available. For most definitive uses— such as index reweighting, updating of item lists, recalculating family budgets or defining criteria for income maintenance payments— there is usually, in any event, a significant timing gap between data availability and action. A more important consideration may be the adequacy of the prior year's data obtained in this manner. These data do not contain the "bounding" safeguards that apply to the current 3 In a continuing operation, the same families would be interviewed over a period of several quarters extending from one year into the next; therefore, the samples for two successive years would not be indepen- dent. The estimation procedure for combining the estimates under these circumstances would be more complex than if discrete samples are used each year, but it wouid still be manageable. year's inquiry; other slippages are also possible— loss, theft, or destruction of items acquired during the previous year but no longer present when the "inventory" is taken. It should be noted that because the prior year's estimates are based entirely on the current inventory of items on hand, they represent the second, estimation method shown in table 6-1, that is, purchases for the family's own use plus the value of gifts received from others. Table 6-3 presents two sets of estimates for 1972, one set based on the survey for that year and the other on prior-year data derived from the survey for 1973. 4 The differences are statisti- cally significant for only about half of the categories shown, but for each of these the prior-year estimate is the higher one. Also, both for major appliances and minor appliances as a group, the prior-year estimate is significantly greater. This pattern suggests that there may be some element of telescoping in the prior-year figures. At the same time, comparison with the corresponding PCE estimates does not reveal a clearly supe- rior match on one basis as opposed to the other. In fact, a merg- er or combination of the two could well achieve the smaller margin of bias as well as the lower sampling error. 4 For purposes of greater comparability, the current year's estimates in this table are also based on the second kind of combination— purchases for personal use plus value of gifts received. 26 Table 6-1.— Aggregate Expenditures for New Appliances: Quarterly Panel Estimates (Two Methods) Compared to Personal Consumption Expenditures, 1972 Category Quarterly panel estimates (millions of dollars) 1 Method 1 Purchases for own use and for gifts given to persons outside the household Method 2 Purchases for own use plus value of gifts received from persons outside the household Ratio of quarterly panel estimates to estimated Personal Consumption Expenditures (PCE) [GNP accounts] 2 Method 1 Method 2 Estimated coefficient of variation of quarterly panel estimates 3 (either method) (percent) Major appliances— Total Cooking stoves and ranges .... Refrigerators and freezers .... Clothes washers and dryers . . . Dishwashers and disposals .... Vacuum cleaners and other floor cleaning equipment Sewing machines and cabinets . Room airconditioners Television sets Black and white Color Yard machinery Minor appliances 4 —Total Small kitchen appliances Portable heating and cooling equipment Electric personal care equipment . . . . , Sound equipment , Photographic equipment Fower tools $10,238.7 978.0 1,636.1 1,453.4 546.9 587.4 404.1 587.4 535.6 2,866.7 643.1 4,305.1 587.9 237.9 322.5 2,260.9 541.5 354.4 $10,691.7 1,005.8 1,707.8 1,513.1 563.0 616.0 437.3 605.1 588.5 3,012.4 642.7 4,767.4 779.9 243.9 368.3 2,417.3 584.4 373.6 .895 .874 .879 .944 1.098 1.013 1.293 1.105 .722 .816 .865 .868 .661 1.135 .955 .974 .636 1.008 .935 .898 .917 .983 1.130 1.063 1.399 1.138 .793 .858 .865 .961 .877 1.164 1.091 1.042 .687 1.063 2.0 5.7 4.9 4.6 6.7 5.1 8.0 5.0 4.8 3.6 7.3 2.4 2.6 5.0 2.8 3.3 6.4 6.1 ^hese estimates are based on special reweighted tabulations from the original survey tapes. They do not reflect editing changes which may have been made by the Bureau of Labor Statistics at later stages of processing and may, therefore, differ somewhat from estimates which have already or will subsequently be published by that agency. In addition to direct purchases and gifts, the estimates under both methods include the value of new appliances 'nstalled in and included in the price of a home. 2 PCE estimates for 1972 used in calculating these ratios were projected by the author from detailed PCE data for 1967 using Census of Manufactures and foreign trade data for 1967 and 1972 (see appendix A). Preliminary. 4 Minor appliance categories include the following: Small kitchen appliances — toasters, mixers and blenders, electric can openers, coffeemakers, broilers, electric pans, electric knives, electric food warmers, electric irons, etc. Portable cooling and heating equipment -portable heaters or stoves, dehumidifiers, humidifiers or vaporizers, electric window or portable fans, etc. Electric personal care equipment— hair dryers, elsctric shavers, electric hair setters, electric toothbrushes, etc. Sound equipment — radios, phonographs, tape recorders, record players, stereo components, etc. Photographic equipment— stil I cameras, slide projectors, movie cameras, movie projectors, etc. Power tools— electric drills, electric saws, electric senders, etc. Table 6-2.— Aggregate Expenditures for New and Used Appliances: Quarterly Panel Compared to Dairy Operation, 1972-73 Average Category Quarterly panel 1 Amount (millions of dollars) Estimated coefficient of variation (percent)* Diary operation Amount (millions of dollars) Estimated coefficient of variation (percent)* * Major appliances 2 Kitchen appliances . . . . Laundry appliances . . . Housecleaning appliances Sewing machines Television sets Minor appliances 3 Small kitchen appliances Sound equipment . . . . $4,265.5 1,773.1 639.0 415.5 3,620.9 599.7 2,475.7 2.5 3.3 3.7 5.7 2.5 1.9 2.4 $2,801.3 1,319.9 445.7 165.6 2,383.4 477.6 1,754.5 10 15 30 35 10 20 10 •Preliminary. ** Rough estimates based on acquisition rates. 1 Method 1 estimates, that is, purchases for own use or for gifts to others. Estimates are 1972-73 averages, whereas those in table 6-1 are for 1972 only. 2 Kitchen appliances Include cooking ranges, refrigerators and freezers, and dishwashers and disposals. Laundry appliances include washing machines and dryers. Housecleaning appliances include vacuum cleaners, eiectric brooms, floor waxers and polishers, and the like. 3 For Inclusions, see footnote 4, table 6-1. 27 Table 6-3.— Expenditures for New Appliances from the Quarterly Panel: Current- Year Compared to Prior- Year Based Estimates, 1972 Category Current year based estimates 1 (1972 estimates based on 1972 survey) Amount (millions of dollars) Ratio to PCE estimate Prior year based estimates 1 (1972 estimates derived from 1973 survey) Amount (millions of dollars) Ratio to PCE estimate Estimated standard error of the differ m. between the two estimates (millions of dollars) 3 Major appliances — Total Ccoking stoves and ranges .... Refrigerators and freezers .... Clothes washers and dryers . . . Dishwashers and disposals .... Vacuum cleaners and other floor cleaning equipment Sewing machines and cabinets . Room air conditioners Television Black and white Color Yard machinery Minor appliances 2 —Total Small kitchen appliances Portable heating and cooling equipment Electric personal care equipment .... Sound equipment Photographic equipment Power tools $10,691.7 .005.8 ,707.8 ,513.1 563.0 616.0 437.3 605.1 588.5 ,012.4 642.7 3,i 4,764.4 779.9 243.9 368.3 2,417.3 584.4 373.6 .935 .898 .917 .983 1.130 1.063 1.399 1.138 .793 .858 .865 .961 .877 1.164 1.091 1.042 .687 1.063 $11,974.4 1,067.6 2,057.1 1.777.7 652.1 771.8 493.9 614.5 656.4 2,888.5 994.8 5,228.2 882.1 271.9 423.7 2,669.5 638.9 342.1 1.047 .953 1.104 1.155 1.309 1.332 1.581 1.157 .885 .823 1.338 1.054 .992 1.297 1.255 1.150 .751 .973 $323 85 131 107 60 50 53 44 42 150 85 170 31 18 16 119 55 31 1 Method 2 estimates, purchases for own use plus value of gifts received from others. 2 For inclusions, see footnote 4, table 6-1 . 3 Preliminary. CHAPTER 7 Expenditures for Home Furnishings and Related Products This chapter embraces a wide range of household furnishings and related commodities not previously covered. Included are furniture, household linens, dinnerware or cookware, decorative items, nonpowered tools, and various other items. Expenditures for these items totaled about $25 to $30 billion annually at the time of the survey. Because of the diversity of items, the survey procedures differed somewhat for the various categories. Most were covered in the quarterly panel, and the information was collected each quarter for some items (mainly household linens) and semiannually for the remainder. Unlike the procedure for household appliances, direct questions were asked about expenditures in the previous 3-month or 6-month periods rather than starting with an inven- tory of items on hand. 1 As usual, all of these categories were also covered in the diary, in which there was a small section set aside for "housewares, furnishings, hardware, and garden supplies." The margin of difference between the survey and independent estimates depends, to some extent, on which source is consulted. It may, therefore, be useful to start with some assessment of the independent data. For the first three categories in table 7-1 — furniture, floor coverings, and household linens— the PCE esti- mates exceed those from the other independent sources. In the light of the experience with certain other categories, it is likely that the PCE estimates will be adjusted downward when the current benchmark revision (to 1972 Census and other data) is completed. The PCE estimate for luggage and related items may also be overstated, perhaps by as much as 15 to 20 percent, according to evidence from the author's work in updating earlier benchmark data. A smailer overstatement, percentagewise, is likely in the PCE estimate shown for dinnerware, glassware, and related items. 2 On the other hand, there is a possibility of understatement in the PCE estimate for window and furniture coverings because of the difficulty of deriving a reliable figure from production data (the usual source for PCE purposes) in a sector in which much of the final output is custom made. A 'The large variety of items of this kind possessed by most households made it difficult to apply the inventory approach in this case. 2 A downward adjustment of some $700 million was made by the author in the Commerce Department's estimates for 1972-73 for this category, to exclude plastic products and a few other items not included in this group in the classification system used in the survey. Some ad- ditional items, possibly representing another $200 million or so in ex- penditures, should probably have been deducted; but the identification of these items was not sufficiently precise for purposes of a further adjustment. 28 similar downward bias probably exists in the estimate for lamps, mirrors, and decorative items, a bias arising from the problems in adequately reflecting expenditures for original paintings and sculptures. Even after allowing for these possibilities, the only category that exhibited a reasonably close correspondence between the survey and independent estimates is the furniture group. This result should not be too surprising in view of the rather consistent finding that higher cost products are more completely reported. Also not unexpectedly, the largest disparities were found for those categories composed of a wide spectrum of relatively small items— dinnerware, glassware, and kitchenware; lamps, mirrors, and decorative items; luggage and related items; and hand tools. The fact that the quarterly panel required a 6-month recall for those categories probably accentuated the problem. The remaining categories— floor coverings, linens, and related items, and window and furniture coverings— seemed to fall somewhere between the two extremes. These categories contain a number of items of sizable cost and/or relatively conspicuous use. Also, for the lower cost categories— household linens and sewing materials— the questioning was carried out on a quarterly rather than a semiannual basis in the interview panel. The independent estimates are too uncertain to permit more detailed comparisons than those shown in the table. Some very rough data suggest, however, that where the survey results were relatively close to the PCE levels (e.g., furniture), the similarity appeared to extend to the various subcategories as well. At the same time, where differences were large (e.g., dinnerware, cookware, etc.), the disparities were widespread over a whole range of items. Timing Differences One of the factors that probably contributed to the under- reporting of expenditures in the quarterly panel was the length of the recall period. Some of the smallest items (household linens and sewing materials) were canvassed each quarter, but the remainder were covered only at semiannual intervals. Table 7-2 provides some indication of the effect of using shorter recall periods for household linens and sewing materials. For comparison with the original quarterly panel data (based on 3-month recall periods), the table provides estimates derived from reported expenditures in only the most recent (final) month of each 3-month period. Because the quarterly panel was 29 divided into three equal subpanels, each operating on a different quarterly cycle, the alternate estimates cover all calendar months of the year and are unbiased from the standpoint of seasonal variation. As shown in the table, the use of data for the most recent month appreciably boosts the survey expenditure levels for both categories. On this basis, the quarterly panel estimate for house- hold linens now exceeds that from the diary operation (table 7-1) and represents 85 to 90 percent of the smaller of the in- dependent estimates; it is still considerably short of the PCE figure, however. For sewing materials, the alternate estimate now just about equals the diary level but remains well below the only available independent estimate from the Census of Business. From the discussion in earlier chapters, it will be recognized that at least part of this apparent improvement is a consequence of "telescoping," that is, in this instance, reporting a purchase as having occurred in the most recent month of a quarter that actually took place eariier. For smaller expenditures of this kind, exaggerations resulting from telescoping are frequently offset, or even outweighed, by the opposite tendency— respon- dents overlooking and omitting some of the purchases actually made during the reference period. As noted previously, these kinds of offsetting errors are present in nearly all survey opera- tions of this nature and have much the same effect as a sampling variance. Because the alternate (most recent month) estimates correspond more closely to, but still do not exceed, the inde- pendent figures, it is likely that they represent the more accu- rate levels, in spite of the telescoping. furniture, the alternate estimate exceeds the independent levels by a moderate margin. Since a high-cost category of this kind is probably more subject to telescoping than to underreporting of purchases due to memory loss, it is likely that shortening the recall period without instituting strong "bounding" controls against telescoping could result in an overstatement of expendi- tures. For two of the other categories— floor coverings and window and furniture coverings— the alternate estimates now begin to approach the independent levels, especially if we accept the lower of the two figures or even some point in between. Al- though these second-quarter estimates are subject to upward seasonal biases for the reasons given earlier, the effect should be somewhat less for these categories than for most of the others. 4 However, as in the case of furniture, telescoping may be more of a factor than memory loss for items of this kind and would add to the upward bias of the alternate estimates. Nevertheless, the increases in the survey estimates arising from the reduced time reference are so large that they suggest a step in that direction could achieve significant gains in accuracy. For the remaining three categories— dinnerware and related items, decorative items, and luggage— it appears that, even if the alternate estimates are accepted at face value, a shortening of the recall period in itself is not likely to overcome the marked reporting deficiencies. In fact, the "second quarter" estimates for these categories are among those most affected by the upward seasonal biases cited above, so that the computed gains may be more apparent than real. An illustration of the use of shorter time periods for the other categories of household furnishings is presented in table 7-3. For these categories— which were canvassed semiannually in the interview panel— the alternate estimates are based on reported purchases in the second (most recent) quarter of each 6-month period. Unlike the items reported on a quarterly basis, the timing of the panel interviews does not permit developing alternate estimates for these semiannual categories which are completely unbiased from a seasonal standpoint. In fact, the "second-quarter" estimates are exaggerated seasonally because the high-purchase month of December is overrepresent- ed in the compilation. 3 This problem is somewhat greater for categories in which gifts are relatively important, such as dinner- ware and similar products and decorative items. The available data, unfortunately, are insufficient for the purpose of con- structing and applying seasonal adjustment factors. In spite of these limitations, the comparisons may cast some light on whether a reduction in the time reference is likely to achieve appreciable gains in reporting accuracy. As might be expected, the second-quarter estimates are substantially higher than the original survey figures in every category. In fact, for Comparisons With Independent Estimates Table 7-1 presents a comparison of the survey results with various independent data for broad categories of items. The independent sources will be familiar from the previous discus- sion. As usual, greatest dependence is placed on the Personal Consumption Expenditures (PCE) estimates prepared in con- junction with the Gross National Product accounts. Also used, in some instances, are the merchandise line data from the 1972 Census of Business and consumer panel estimates from the Market Research Corporation of America adjusted to client production and sales data. In view of the complexity of the categories covered, the caveats cited in previous chapters con- cerning the precision of these independent estimates apply with equal or greater force in this case. Moreover, there is consider- able divergence among the sources in several instances, which accentuates the need for caution. One of the more surprising findings exhibited in the table is the similarst/ between the quarterly panel and diary results for 3 This tendency may be slightly offset by the overrepresentation of some of the low-purchase summer months in the second-quarter esti- mates as well. 4 December is not as much of a seasonal peak for purchases of these items (or for furniture) as it is for some of the other categories or other kinds of merchandise. 30 most categories. The original intention was to depend principal- ly on the interview panel for home furnishings and related expenditures, mainly because the relatively high sampling variances for certain of the categories could preclude the use of the diary estimates, except for very general purposes." However, variances aside, the most striking difference between the two sets— that for dinnerware, glassware, and related items— was clearly in favor of the diary. Apparently, the 6-month recall period used in the quarterly panel was far too long to obtain even a semblance of complete reporting for this highly diversi- fied category, which included many small, infrequently pur- chased items. Table 7-1.— Aggregate Expenditures for Household Furnishings and Equipment: Quarterly Panel and Diary Operation Compared to Independent Sources, 1972-73 Category Survey estimates 1 Quarterly panel Amount (millions of dollars) Estimated coefficient of variation (percent) 2 Diary operation Amount (millions of dollars) Estimated coefficient of variation (percent) 2 Ratio of survey estimates to Personal Consumption Expenditures (PCE) IGNP accounts] 3 Quarterly panel Diary operation Census of Business Merchandise line data Quarterly panel Diary operation Adjusted National Consumer Panel Data (Market Research Corporation of America) 4 Quarterly panel Diary operation Furniture* Floor coverings* Soft surface Hard surface Household linens Sewing materials Window and furniture coverings Dinnerware, glassware and kitchenware* Lamps, mirrors, clocks & decorative items . . . . Luggage and closet storage items* Hand tools (except garden tools) $9,494 2.7 3,136 3.2 2,860 3.2 276 — 1,907 3.0 1,878 3.0 1,554 3,5 918 3.2 1,200 4.1 278 .5.1 $9,390 2,350 2,080 270 1,934 2,292 1,444 1,964 1,034 322 339 10 7 7 7 5 10 5 10 10 10 .899 .645 .635 .782 .576 ,800 .226 .567 .285 .889 .484 .462 .765 .584 .744 .484 .488 .330 .579 .926 .795 .862 .790 .579 .632 .916 .595 .629 .801 .707 .588 .860 .736 .625 .747 •Survey and PCE estimates for these categories are 1972-73 averages. All other estimates in table are for 1972 except Market Research Corporation of America data which are for 1973. 1 These data are derived from reweighted original data tapes and do not incorporate editing changes which may have been made at later stages of proces- sing by the Bureau of Labor Statistics. The estimates may differ somewhat, therefore, from those already published or to be published by that Agency. Whereas quarterly panel data are for calendar years, diary estimates are for fiscal years, July 1972-June 1974. 2 Rough preliminary estimates. Sample sizes used in these computations are roughly twice as large for those categories denoted by asterisks (*) in the first column as for the remaining categories. 3 For categories denoted by asterisks ( * ) in the first column of the table, the PCE data used to compute these ratios represent current estimates of the Department of Commerce for the appropriate years (not yet adjusted to the most recent 1972 benchmarks). The Commerce estimate for "dinnerware, glassware, etc." has been adjusted downward by the author by some $700 million to eliminate plastic products and a few other items not included in this category in the classification system used in the expenditure survey. For categories not marked with asterisks, current PCE estimates were prepared by the author from PCE benchmark data for 1967 using Census production and foreign trade data for 1967 and 1972 (See appendix A). 4 Computed from basic panel data using adjustment factors supplied by MRCA (See appendix A). Table 7-2. -Aggregate Expenditures for Household Linens and Sewing Materials: Quarterly Panel Estimates Based on Data for All Months and on Reported Purchases in Final Month of Each 3-month Period, Compared to Independent Data Sources, 1972 [Millions of dollars] Quarterly panel estimates Ratio of quarterly panel estimates to: Category Based on data for all months 1 Based on reported purchases in final month of each 3-month period Personal Consumption Expenditures (PCE) [GNP accounts] Census of Business Merchandise line data Data for all months from panel "Final month" data from panel Data for all months from panel "Final month" data from panel $1,907 $2,191 1 878 OtKO .576 .662 .791 .908 579 704 'Data collected each quarter based on reported purchases in previous 3 months. See table 7-1 for estimated coefficients of variation. 31 Table 7-3.— Aggregate Expenditures for Selected Household Furnishings: Quarterly Panel Estimates Based on Data for All Months and on Reported Purchases in Final 3 Months of Each 6-Month Period, Compared to Independent Data Sources, 1972-73 r »»,i- * ■ ,, i [Millions of dollars] Quarterly panel estimates Ratio of quarterly panel estimates to: Category Based on data for all months' Based on reported purchases In final 3 months of each 6-month period Personal Consumption Expenditures (PCE) [GNP accounts] Census of Business Merchandise line data Data for all months from panel "2nd quarter" data from panel Data for all months from panel "2nd quarter" data from panel $9,494 $11,319 3.136 3,764 1.554 1,938 918 1,276 1,200 1.606 .899 1.071 .645 .774 .800 .998 .226 .314 .567 .759 .926 1.104 Floor coverlnas,* new Window and furniture Dinnerware, glassware. .795 .954 .632 .789 Lamps, mirrors, & decorative items Luggage & related items* 278 387 .285 .397 - - •Survey and PCE estimates in this table are 1972-73 averages. All other estimates are for 1972. 1 Data collected at 6-month intervals based on reported purchases in previous 6-months. See table 7-1 for estimated coefficients of variation. CHAPTER 8 Automobile and Other Vehicle Expenses This chapter is concerned with some of the largest items in the American family budget— the purchase, maintenance, and operation of automobiles and other vehicles. Expenditures for these purposes aggregated over $100 billion annually at the time of the survey. Because of the diversity of products and services, both the quarterly panel and the diary operation were enlisted in this cause. The interview panel was almost automatically assumed to be the source for data on vehicle purchases, in view of the large sampling variances associated with this item. How- ever, there was much less certainty concerning the optimal source for the remaining categories. Detailed questions were included each quarter in the interview panel covering expendi- tures for maintenance and operation of automobiles and other vehicles during the preceding 3 months. These items were also specifically covered in the diary operation in a small section set aside for "gas, oil, tolls, parking fees, and other vehicle expenses." Automobile Purchase The "inventory" approach was used in the quarterly panel to determine expenditures for automobiles and other vehicles. At the first quarterly visit, questions were asked about possession of vehicles of all kinds, and various details about those vehicles on hand were requested, including a description of each and the date of acquisition. Respondents were also asked whether any additional vehicles had been acquired during the year, and, if so, at what price. The procedure was not quite as precise as that used for household appliances, but the fact that automobiles constitute the largest and most costly class of durables probably means that less probing would ordinarily be required to obtain results of comparable reliability. Table 8-1 compares the survey estimates of expenditures for new and used cars with those from various independent sources. 1 In addition to two familiar sources— the Personal Consumption Expenditures (PCE) estimates and the merchandise line data from the 1972 Census of Business— a third source was used: the Annual Housing Survey conducted by the Bureau of the Census, which collects information on the housing characteris- tics of the population and, for certain years, data on expendi- tures for selected appliances and equipment items. Three different survey estimates are provided in the table in order to achieve greater comparability with the independent data. The gross value (before trade-in) and the net value (after trade-in) estimates are probably reasonably comparable with the corresponding Annual Housing Survey figures. Both these sources include not only automobiles purchased for personal (nonbusiness) use, but also those possibly used partly for business. Those purchased exclusively for business use, however, would be excluded. As is evident, the expenditure survey and housing survey estimates are very close, for both new and used cars and on both a gross and net basis. The survey estimate that is probably closest to the PCE level for new cars is the one based on gross value after adjustment for business use. 2 For vehicles purchased partly for business use, survey respondents were asked for the estimated percentage of mileage that would be devoted to business purposes. This estimate was the principal basis for adjusting the survey figure. On this adjusted basis, the survey and PCE estimates for new cars correspond closely in both years. Certain conceptual dif- ferences remain which could affect this comparison. For exam- ple, the PCE estimate includes expenditures made by nonprofit institutions, probably a very small percentage of the total. On the other hand, retail sales discounts are presumably excluded from the PCE estimates but may be included, in part, in the "gross" survey estimates. It is not likely, however, that these kinds of differences could markedly reduce the correspondence exhibited in the table. The Census of Business figure includes retail sales of automo- biles both for business use exclusively and for other purposes, and this undoubtedly accounts for the higher expenditure level shown for new cars. Although the disparity between the Busi- ness Census and the survey figures is difficult to assess, it does not appear to be excessive in view of this difference in coverage. The Business Census estimate for used cars is undoubtedly understated; many of these are sold directly by private owners rather than through sales outlets. Comparisons for other vehicle categories are complicated by classification problems 3 and by more widespread use for busi- ness purposes. Because of classification difficulties, for example. Only the quarterly panel estimate? are being used in this comparison. The diary-based estimates were only about half as large, and this fact confirmed previous findings that short-term recordkeeping is generally inadequate for measuring large expenditures of this kind. The PCE estimate for used cars represents only the profit margin for sales of used cars plus the value of used cars transferred to private use from Government or business use. This figure is, therefore, not used in these comparisons. 3 For one important category- motorcycles and scooters (for which the survey estimates were close to $1 billion annually)— the PCE estimate is included as part of a much larger aggregate so that no direct compar- ison was possible. 32 33 a comparison between the survey and PCE data is possible only for new trucks, trailers, and recreational vehicles combined. On this basis, the survey figure (after adjustment for business use) amounted to about $4.9 billion annually, which was about 10 percent below the PCE estimate. However, uncertainty about the accuracy of the business-use allocation in PCE could mean that the true difference was either smaller or larger. For boats, the survey estimate of some $850 million annually (for new boats) was some 30 percent under the PCE aggregate. This dif- ference could also be affected by the "business-use" allocation, but population coverage could be a minor contributing factor. Some people purchase boats for use as their main living quarters, but such acquisitions might be largely omitted in a household survey because of the difficulty of locating and covering that segment of the population. Gasoline and Oi! Comparisons between the survey and the independent estimates of expenditures for gasoline and oil are provided in table 8-2. The estimates relate to expenditures for all vehicles combined, because separate data for automobiles are not available from the independent sources. The diary estimates have been adjusted upward by about $2.3 billion in each year to account for gasoline purchased while on trips or vacations. In the survey procedure, no entries were to be made in diaries during periods in which all family members were away overnight or longer. However, gasoline expenditures might, of course, be especially large on such occasions. The adjustment was based on a special inquiry in the quarterly panel which obtained an itemized account of a family's expenditures for each trip which lasted overnight or longer. After this adjustment, the diary and quarterly panel estimates are not significantly different. Ordinarily, some understatement might be expected in the diary estimates because the home- maker, who was usually the recordkeeper for the family, might not be aware of all of the gasoline purchases made by the hus- band or the young drivers in the family. However, gasoline is one of the most regularly purchased items, and the home- maker might be more inclined than in othei cases to inquire about such outlays when completing the record. As the table indictates, the survey estimates are quite close to both sets of independent data. One reason they might slightly exceed the PCE levels is that they probably include a small mar- gin of expenditures for business use which are presumably excluded from PCE. The Census of Business estimates, on the other hand, probably reflect an appreciable amount of sales for business purposes but may exclude some sales for private use which are made in repair shops or other establishments outside of the coverage of that Census. Vehicle Repair, Maintenance, and Miscellaneous Expenses Table 8-3 presents a comparison between the survey and inde- pendent estimates for the remaining major categories of vehicle expenses— those for tires and accessories, maintenance and repair, licensing, and insurance. There was considerable un- certainty during the planning stages of the survey as to whether the quarterly panel or the diary operation would be the more effective instrument for these purposes. As demonstrated in the table, the quarterly panel appears to be considerably superior to the diary in most categories. Moreover, there is a high degree of correspondence between the interview panel and the independent estimates in most cases in which comparisons are possible. This outcome was somewhat unex- pected for the broad category of repair and maintenance expen- ditures where a good deal of difficulty had been anticipated in obtaining accurate measurements. Of course, the independent PCE estimate may also be subject to a good deal of inaccuracy in this particular instance, and the lack of itemization in the comparisons could conceal some appreciable differences. The close match for vehicle insurance premiums, like that for health insurance (see chapter 10), suggests that the modified "inventory" approach 4 used for insurance in general was effec- tive. The disparity for vehicle accessories indicates, not surpris- ingly, that a 3-month recali period was probably too long for a category consisting of a wide diversity of relatively small products. 5 Although the data from the diary operation appear to be defi- cient in many respects, there are some insights into the reasons for these differences. One apparent problem, also observed in previous reports, was that many diary entries were incomplete, probably because of insufficient space for the expenditure category and/or inadequate attention by interviewers to this matter at the time of diary pickup. As a consequence, there is a large residual expenditure category ("combined or un- specified costs") which obscures some of the comparisons. If this amount were distributed among the specific categories it might eliminate much of the gap for vehicle repairs and at least reduce the differences for other items. 6 Another likely explanation for the apparently understated diary levels for accessories, tune-ups, and such items is one al- ready mentioned in the discussion about gasoline, that is, that a single respondent (generally the homemaker) kept the diary for the entire family. Maintenance of vehicles is,of course, one responsibility most likely to be assumed by the husband or. The procedure called for recording, at the first interview, all of the policies in effect at that time, together with a description of each. At the final interview a year later, information on premiums paid during the year was obtained for those policies still in effect. Ne-vly acquired policies were also recorded at that time, together with the premiums paid up to that point. Alternate estimates based on reported expenditures in the most recent month of each 3-month period showed a higher level of expen- ditures for accessories of about 10 percent, still a large deficiency in comparison to the PCE level. 'This residua! category probably includes some expenditures for gaso- line made in combination with other outlays and some miscellaneous items, so that the amount allocable to lepairs and maintenance is un- certain. 34 for second or third cars, by the younger drivers as well. Because such expenditures are irregular in nature, the homemaker might be less likely to think of and ask about them than in the case of gasoline. Although the quarterly interviews are also usually conducted with the homemaker, these panels allow more time and opportunity than do diary records for the respondent. through observation, family discussions, receipt of bills, and the like, to become aware of expenditures made by others. At the same time, the relatively low diary estimates for vehicle regis- tration and insurance premiums confirm another finding-that a short-term recordkeeping approach is often ineffective for very infrequent types of expenditures. Table 8-1. -Aggregate Annual Expenditures for New and Used Automobiles: Survey Estimates (Quarterly Panel) Compared to Various Independent Sources, 1972-73 Source Survey estimates (quarterly panel) 1 Gross value (before trade-in allowance) Net value (after trade-in allowance) Gross value after adjustment for business use-* Personal Consumption Expen- ditures (PCE! (GNP Accounts). . Census of Business Merchandise Line Data Annual Housing Survey (Bureau of the Census) Gross value (before trade-in) Net value (after trade-in) . . . New automobiles Amount (billions of dollars) 1972 1973 $34.7 28.8 32.2 32.1 36.8 $36.6 30.2 34.4 34.5 37.1 30.7 Estimated coefficient of variation (percent) 2 3.5 3.5 3.5 2.0 2.0 Used automobiles Amount (billions of dollars) 1972 1973 $19.0 16.8 18.1 14.1 $18.9 17.2 17.9 Estimated coefficient of variation (percent) 2 3.5 3.5 3.5 18.4 16.5 2.0 2.0 Table 8-2.-Aggregate Annual Expenditures for Gasoline and Oil: Survey Estimates Compared to Independent Sources, 1972-73 Source Survey Estimates 2 Quarterly panel Diary operation Personal Consumption Expenditures (PCE) (GNP Accounts) Census of Business Merchandise Line Data Amount of expenditure (billions of dollars) 1972 1973 $25.4 3 24.0 24 9 26.3 $29.1 3 28.1 27.8 Estimated coefficient of variation (percent) 1 , 1.5 2.5 Preliminary. .'«^7 : ;££*«;^^^ panel on itemized expenses incurred on trips Table 8-3.— Aggregate Annual Expenditures for Vehicle Repair, Maintenance, and Miscellaneous Items: Survey Estimates Compared to Independent Sources, 1972-73 35 Survey estimates Ratio of surve y estimates to Quarterly panel Diary operation Expenditure (PCE) estimates Category Amount per year (billions of dollars) Estimated coefficient of variation j (percent) Amount per year (billions of dollars) Estimated coefficient of variation (percent) 1 Quarterly panel Diary operation Tires and accessories Tires (except recaps) $4.0 2.0 1.2 1.5 13.1 1.5 8.8 2.0 2.8 0.8 0.7 3.9 (NA) 13.9 10 $2.9 7.0 1.2 5.0 11.9 5.0 7.8 5.0 0.6 0.6 0.4 22.5 1.6 (NA) 10.1 (NA) .938 .676 .673 .672 Repair, maintenance, and Miscellaneous expenses Total, except registration, licensing, and insurance . . Vehicle repairs ..... Tune-ups, lubrication. . .892 .811 Vehicle rental Combined or unspecified costs . . . Vehicle registration, licenses, inspection, and docking fees Vehicle insurance (NA) (NA) 3 927 3 fiOfi (NA) = Not available. ' Preliminary. 2 Consists primarily of cases where diary entries were not sufficiently precise to allocate the expenditures to any of the categories shown. Also includes a small amount ($150 million) for car washing. May also include some purchases of gasoline made in combination with other expenses. 3 Because PCE data on insurance premiums have a different conceptual basis (see appendix A), the comparison here uses insurance industry figures instead. The source is aggregate premium data for private vehicles compiled by A.M. Best and Co. The ratios in the table are based on a comparison of the survey and A.M. Best figures for 1973. CHAPTER 9 Housing Costs This chapter deals with housing costs, which have replaced food expenditures as the largest single element in the American family budget. The topics covered here include rent, mortgage payments, real estate taxes, fuel and utility costs, and property maintenance and improvement. Although the quarterly panel was regarded as the principal source of this information, a parallel set of estimates can be constructed from the diary results. Survey Procedures for Housing Items A variety of rather complex procedures was used in the quarter- ly panel for purposes of measuring housing expenditures. At the first interview, an "inventory" was recorded of all real property owned or being bought by the respondents. The inventory de- scribed the type of property; whether it was for residential, business, or rental use; the date of acquisition; and, if the prop- erty was obtained within the previous 3 years, its price and closing costs. 1 At the final visit a year later, this inventory was updated to record any newly acquired property and to identify any property sold or otherwise disposed of during the year. For families who had moved into sample units during the course of the year, the initial inventory also included property owned at any time during the prior year. For each owned property in the "inventory," information was obtained at the final quarterly visit concerning indebtedness, including a good deal of information (size of debt, interest rate, duration, monthly payment, etc.) about each outstanding mortgage. In addition, sufficient information on number of payments, special fees, and other mortgage-related expenses, was recorded to derive total annual outlays. At this final inter- view, other questions were addressed to related property costs, such as real estate taxes, special assessments, and ground rent. For renters, information was obtained at the first visit on the monthly (or other periodic) rent, and the kinds of facilities and services, if any, included in the payment. The same type of information was recorded at the final visit in order to identify any changes that may have occurred. These data, together with information on prior rental quarters for in-movers during the year, provided the basis for calculating annual expenditures for rent. At each quarterly interview, an inquiry was made on expendi- tures for repairs, alterations, and maintenance of owned proper- ty. 2 The procedure called for identifying any projects of this kind undertaken during the previous 3 months and for detailing of any contract costs or direct expenditures for material and supplies. A special technique, known as the "last payment" principle, was used each quarter to determine utility costs. 3 Instead of asking directly about expenditures during the previous 3 months, the interviewer asked for the amount of the most recent bill and for the period (1 month, 2 months, etc.) to which it related. These data provided a basis for converting the information to a consistent time period. For fuel expenditures, however, direct questions were asked about purchases in the preceding 3 months. Only minimal provision was made in the diary operation for housing expenditures. A small section was set aside on the diary record for "rent, utilities, fuel, phone, insurance." Comparisons With independent Data Table 9-1 compares the survey results for housing expenditures with various independent data. A main independent source, as for other categories, is the Personal Consumption Expenditures (PCE) estimates. No comparison with PCE is possible, however, for some of the largest segments of housing costs-mortgage finance charges, property taxes and assessments, or home repairs or alterations. For PCE purposes, an artificial measure, theoreti- cally representing the rental value of owner-occupied properties, is substituted for these kinds of ownership costs. 4 A few com- parisons are made with data from two Census Bureau sample surveys, the Annual Housing Survey (AHS) and the Survey of Residential Alterations and Repairs (SORAR), which are described in appendix A. Continuing the somewhat surprising pattern exhibited for various other expenditure classes, there appears to be a good deal of similarity between the quarterly panel and diary esti- mates for several categories of housing costs, especially for rent, utilities, and fuel purchases. In the diary operation, the exis- 1 At the second quarterly visit, considerable detail was also obtained about the living quarters currently occupied by the family (e.g., number 3f rooms, type of heating system and fuel used, age of structure, etc.). Fhis detail was obtained for rental as well as for owner-occupied units. 36 2 The questions were adapted from the Census Bureau's Survey of Residential Alterations and Repairs. 3 This technique was adapted from the continuing consumer expendi- ture survey in the United Kingdom. 4 Interestingly, the PCE annual aggregate estimate for this item (about $80 billion at the time of the survey) was very close to the results of a special question included in the expenditure survey on what homeowners believed their homes would rent for on the open market. 37 tence of a large residual group under utilities, resulting mainly from incomplete or inadequately described entries in the record book, obscures some of the more detailed comparisons but does not detract from the essential pattern. The much lower diary estimate for home-mortgage payments might be attributable, in part, to the fact that some such dis- bursements are made directly from bank accounts and could be forgotten by respondents. Also, some mortgages are repaid on an irregular basis and this irregular payment could cause some slippage in the diary recording. No direct comparison is possible with the data presently available for home repairs and altera- tions, because the quarterly panel estimates relate to all real property owned by the family, whereas the diary estimates theoretically exclude property used for business purposes or rented to others. 5 Data from SORAR suggest, however, that the difference between the quarterly panel and the diary esti- mates is not inconsistent with that expected from the difference in coverage. Comparison of the surveys with the independent sources reveals a considerable degree of correspondence, especially for the quarterly panel estimates. In fact, many of the ratios are not significantly different from unity, which of course would constitute an exact match. This relative success in measuring housing expenditures was not entirely unexpected in the light of previous experience in this field. 6 The regularity of many of these outlays is undoubediy a contributing factor; the introduc- tion of special techniques in the interview panel, such as record- ing an inventory of ali properties at the outset and using the so-called "last payment" principle for utilities, also added some refinement to the methods used heretofore. 5 Comparisons on a more consistent basis should be possible from data which will be available in the future. 6 See, for example, Peter Frontzak and David Koons, "Study of Home- makers' and lenders' Responses for Monthly Mortgage Payments, Yearly Real Estate Taxes, and Yearly Property Insurance Payments," Paper presented at American Marketing Association/Bureau of the Census Seminar on Survey Methodology, Washington, October 1976. The only major disparity between independent estimates and both survey estimates is for fuel costs, but part of that differ- ence can be ascribed to conceptual factors. The PCE estimate includes all fuel used for rental quarters, regardless of whether the landlord or the tenant was responsible for providing heating and hot water. The survey estimates, on the other hand, include fuel purchases for rented quarters only if the tenant paid directly for those services (mainly for rented one-family homes and some older multifamily units). The amount attributable to this conceptual difference cannot be ascertained with any preci- sion, but a "ballpark" estimate would place it at $1 billion at the very minimum. The survey-PCE ratio would increase to somewhere around 0.80 if a minimal allowance of this magni- tude were made. 7 It should be stated that although the overall comparisons for home repairs and alterations were reasonably close, there were a number of differences in detailed components (not shown in the table). For example, the SORAR data indicate that about 40 percent of total expenditures were for repairs and mainte- nance (as opposed to structural changes), whereas the cor- responding quarterly panel figure was only 25 percent. Aside from the higher percentage allocated to structural changes, the quarterly panel data are distributed differently among additions, alterations, and replacements. Detailed comparisons between the diary estimates and SORAR are precluded because of the existence of a large undifferentiated group of repair and replace- ment expenditures in the former. However, the diary estimates appear to show a larger proportion of expenditures for materials and supplies and less for labor and service contracts than does SORAR. Whether these various deviations at detailed levels are attributable to differences in classification, to sampling errors (which are quite substantial for the various components), or to other causes is not known. However, the subject bears further investigation. 7 This type of conceptual difference would also affect the compar- isons for utility costs. If a correction could be made for this factor, the survey-PCE ratios might increase by several percentage points. 38 Table 9-1.— Aggregate Annual Estimates of Housing Expenditures: Quarterly Panel and Diary Operation Compared to Independent Sources, 1972-73 Category Survey estimates 1 Quarterly panel Amount (billions of dollars) Estimated coefficient of variation (percent) 2 Diary operation Amount (billions of dollars) Estimated coefficient of variation (percent) 2 Ratio of survey estimates to independent sources Quar- terly panel Diary oper- ation Source of indepen- dent data 3 Contract rent 4 Mortgage payments 6 Property taxes 7 Home repairs and alterations All owned property Own home* Utilities-total Electricity and gas Water, sewerage and trash collection . . Telephone and telegraph Cable TV Type not specified Fuel (fuel oil, coal, tank gas, etc.) $40.0 52.1 19.1 18.9 (NA) 36.2 19.2 4.2 12.4 0.5 4.4 2.0 3.0 3.5 5.0 1.5 1.5 2.0 1.0 (NA) 2.0 $38.8 35.1 (NA) (NA) 12.5 38.7 17.9 3.3 13.6 0.3 3.6 4.2 3.0 5.0 15.0 2.0 2.0 3.0 2.0 (NA) 5.0 1.043 1.044 .954 1.048 .981 .976 1.055 .931 1.028 .630 1.013 .703 1.037 9 1 .049 .594 5 PCE AHS AHS SORAR SORAR PCE PCE PCE PCE PCE PCE (NA) = Not available. 1 The data are derived from reweighted original data tapes and do not incorporate editing changes which may have been made at later stages of proces- sing by the Bureau of Labor Statistics. The estimates may differ somewhat, therefore, from those already published or to be published by that agency. Preliminary. The coefficients of variation of the two AHS base figures used in computing the ratios of survey to independent estimates were about 2 percent (mortgage payments) and 2.5 percent (property taxes), respectively. The coefficients of variation of the SORAR figures used for home repairs and alterations were about 3 percent in both cases. The PCE estimates are not constructed in a manner which would permit calculation of sampling variances. 3 Independent sources are as follows: PCE— Personal Consumption Expenditures, GNP Accounts AHS — Annual Housing Survey, Bureau of the Census SORAR— Survey of Residential Alterations and Repairs, Bureau of the Census (PCE and SORAR data are 1972-73 averages; AHS data are for 1973.) 4 Includes any utilities or facilities covered as part of the periodic rent payment. Excludes utilities paid separately by consumer unit. 5 PCE estimates adjusted upward by author by about $4.7 billion to restore amounts deducted by Commerce Department in converting census contract rent figure to "space" rent concept used for PCE purposes (see appendix A). includes principal, interest, and, if covered by periodic payment, property taxes and insurance. 7 Includes both property taxes paid as part of periodic mortgage payments and those paid separately by homeowner. There is some duplication, there- fore, between the "mortgage payments" and "property tax" categories in the table. 8 For diary estimates, excludes expenditures for business purposes (presumably on business property or property rented to others). For SORAR esti- mates used to compute ratio, refers to 1- to 4-unlt owner-occupied property. These categories would approximate expenditures on owner-occupied property. 9 Ratios for diary estimates shown only for total utilities because of large residual group (utility type not specified) in distribution by utility type. CHAPTER 10 Health Expenditures This chapter focuses on one of the major problem areas in the expenditure field— the measurement of health expenditures. Aside from the intricacies arising from the diver- sity of services and products this category embraces, reporting of health expenditures is greatly complicated by the pervasive influence of health insurance and other so-called third-party payors. The objective in the 1972-73 survey was to measure out-of-pocket costs, that is, the family's net expense for health services after it has been reimbursed by insurance or any other source. For this purpose, a detailed series of questions was included in the interview panel on a semiannual basis covering the following areas: the use of any kind of health services during the preceding 6 months; the total cost of such services; the actual or anticipated reimbursement, if any, from insurance or other sources; and the net expense to the family. Purchases of a few major health products— prescription drugs and medical appliances— were also coveted by the questioning. As usual, all expenditures were also covered by the diary opera- tion. A small section on the diary form was set aside for "per- sonal care, drugs, and medical supplies." However, the form did not specifically mention the main body of health expenditures (i.e., hospital charges, physician services, etc.), and these were relegated to various catchall sections. Previous Efforts To Measure Health Expenditures Because of the large and rapidly increasing cost of health ser- vices, numerous attempts have been made to measure and chart expenditures in this field. A discussion of these efforts may be helpful both as background and because some of the data will be used as yardsticks in appraising the current survey results. Perhaps the longest standing endeavor of this kind is the prepa- ration by the Social Security Administration (SSA) of annual aggregates of national health expenditures in various broad cate- gories—hospital care, physicians' services, drugs and drug sun- dries, etc. Data are developed on the total cost of such services and products and on the source of funds for payment— whether payment is made directly by the consumer or through private health insurance. Government, philanthropic, or industrial enterprises, and the like. A variety of primary data sources is used in the preparation of these estimates, including informa- tion on health providers from Internal Revenue Service records and statistics compiled by the American Hospital Association and other private health-related organizations. As «s customary for constructed series of this kind, numerous transformations of the original data are made, and there is no way of assessing the validity of the results. A parallel effort is represented by the Personal Consumption Expenditures (PCE) estimates prepared by the Commerce Department in conjunction with the Gross National Product accounts. For health services, the primary data sources used are broadly similar to those consulted by the Social Security Ad- ministration. The PCE estimates, in this case, comprise the total cost of the services; including that part defrayed by health insurance. For health products, the PCE estimates are developed from production data (mainly from the Censuses of Manufac- tures) converted to a purchase value basis. More direct attempts to measure health expenditures have been made through various survey mechanisms. On occasion (most recently for 1970 and 1974), an inquiry on health expenditures has been appended to the continuing Health Interview Survey (HIS) conducted by the Census Bureau under the sponsorship of the National Center for Health Statistics of the U.S. Public Health Service. The procedure in these recent efforts has been to drop off questionnaires on this subject at the conclusion of the regular HIS interview. The questionnaires are self -adminis- tered and mailed in by respondents. A separate form is provided for each family member, requesting estimated total (out-of- pocket) expenditures in the previous year for a few broad cate- gories—dental bills, doctors' bilis, hospital bills, prescription medicines, eyeglasses, and miscellaneous items— and for health insurance premiums. A more painstaking approach is that employed in a series of intermittent household surveys (the most recent covering 1970) by the Center for Health Administration Studies (CHAS) of the University of Chicago. The procedure was to conduct an ex- tremely detailed interview with the family to identify all health- related episodes during the previous year, such as hospitaliza- tions, serious accidents, pregnancies, chronic conditions, etc., which might entail substantia! health expenditures. The inter- viewer also inquired about treatment and costs for each such episode. A final series of questions addressed residual, smaller expenditures for the same period. Various other survey techniques have also been used in this connection from time to time. One technique was to identify all health practitioners used by the family during the reference period and to inquire about treatments and expenditures in- volved in each case. 1 Another technique for recording health ^ent H. Marquis, M Susan Marquis, and Joseph P. Newhouse, "The Measurement of Expenditures for Outpatient Physician and Dental Services: Preliminary Findings from the Health Insurance Study," Pro- ceedings of the American Public Health Association Meetings, Chicago, November 1975. 39 40 • services used was to experiment with diaries for more extended periods, such as over a calendar quarter, with periodic collection in between. 2 The Health Resources Administration of the U.S. Public Health Service, after a period of research and experimentation, is currently sponsoring a new undertaking in this field. The basic procedure is to cover a national household panel over a 12- month cycle at 2-month intervals to measure use of health services and expenditures for this purpose. Personal interviews are being used for the first two intervals and for the final in- quiry. Telephone contacts are employed for the intervening periods. Respondents are asked to keep records of expenditures and related items during each interval, and these records are used for reference purposes in conducting the interviews. Although specific arrangements have been made for oniy one annua! cycle of this kind, the hope is that this new procedure will be the starting point for at least an intermittent series of surveys. 3 Quarterly vs. Diary Results Turning once again to the 1972-73 survey, it may be useful to begin with a comparison between the quarterly panel and the diary results for the various categories. Partly because of sam- pling variances, the initial intention was to depend largely on the panel for the main body of health expenditures. Also, it was believed that an interview procedure might be necessary to determine the impact of health insurance and other third-party payors. The diary, however, was regarded as the likely, if not only, source for some of the smaller expenditures, principally medical supplies and over-the-counter drugs. In spite of expectations to the contrary, table 10-1 reveals a rather surprising similarity between the diary and quarterly results for some of the main categories of out-of-pocket health expenditures. Unfortunately, the coding and classification procedure used in the two operations differed in a number of respects, so only broad comparisons are possible with the data presently available. However, the levels are almost identical for professional medical care and services, dental care, and eye care. The higher level of expenditures the interview panel shows for hospital services may largely reflect the inability to identify from sometimes incomplete diary entries all of the subsidiary services (x-rays, laboratory tests, etc.) that were provided in hospitals. (Such services were assumed to be rendered outside unless specifically indicated as given in hospitals.) 2 Seymour Sudman, Wallace Wilson ^nd Robert Ferber, "Cost Effec- tiveness of Using the Diary as an Instrument of Collecting Health Data in Household Surveys," Preliminary Report to Bureau of Health Ser- vices Research and Evaluation, HRA, Public Health Service, Survey Research Laboratory, University of Illinois, Urbana, 1974. This study dealt with use of health services rather than expenditures, but the tech- niques and findings should be applicable to expenditures as well. 3 The initial survey will be conducted under contract with the Research Triangle Institute and the National Opinion Research Center. The panel will consist of approximately 10,000 households. The obviously understated diary estimate for health insurance premiums is understandable, because so large a proportion of these are paid through payroll deductions and might be un- known to or at least easily overlooked by the family record- keeper. On the other hand, the interview panel seems to have understated the few small expenditure items (e.g., prescription drugs) included on the questionnaire. The fact that a 6-month recall period was used in the panel for all health expenditures undoubtedly contributed to the deficiency. Overall Comparisons With Independent Sources A comparison of the overall survey data with various indepen- dent sources described earlier in this report is presented in table 10-2. The comparisons relate to out-of-pocket health expendi- tures, excluding drugs and health insurance premiums, because these particular aggregates could be derived for ail of the sources. As is evident, the survey estimates are somewhat, but not markedly, below two of the independent sources— the SSA and CHAS estimates. The difference with SSA almost disappears, however, when an adjustment is made to exclude nursing home care, which represents a conceptual difference between the sources. The 1972-73 survey was limited to the civilian noninsti- tutional population, and expenditures for nursing home care would be included only if actually paid for by residents of the survey units. Inclusion of nursing home care could occur if a person who was living in the survey unit at the time of interview had spent some time in a nursing home during the 6-month reference period used in asking about health expenditures. More likely, the survey would cover payments made by family mem- bers on behalf of relatives who were longtime residents of nursing homes. In either case, the total expenditures for these purposes would likely cover only a small fraction of the actual costs of nursing home care in the United States. The largest part of expenditures undoubtedly are made by the nursing home residents themselves, by public or private agencies, or by third- party payors. The SSA estimates cover the entire population, including residents of nursing homes, and would include pay- ments made by such persons directly as well as those made on their behalf. In contrast, the comparison between the 1972-73 survey and CHAS estimates should not be materially affected by this factor, because both covered only the civilian noninsti- tutional population. Therefore, an adjustment in CHAS figures to exclude nursing home care (which cannot be made with available data) would not be likely to reduce the gap to any extent. 4 As is also apparent, the third independent source in table 10-2— the HIS estimate-exceeds the 1972-73 survey levels by a sizable margin. There is reason to believe, however, that the procedure followed in HIS— using self-administered questionnaires to obtain annual estimates for very broad categories of services- can result in an exaggeration of expenditures. In fact, previous "However, the differences between the survey estimates and those of CHAS were just about on the borderline of statistical significance before adjustment. ■ I analyses of HIS estimates reveal that they consistently exceed the levels in SSA and other independent sources, often by appreciable amounts. 5 The differences shown here, therefore, cannot by themselves be construed as indicative of serious deficiencies in the expenditure survey results. It may be noted that, for the first time in this report, the GNP Personal Consumption Expenditures (PCE) estimates are not used as a principal basis for comparison in this case, although some limited use is made of them in examining drug expendi- tures. However, as previously indicated, the PCE estimates for health services in general incorporate the contribution from health insurance and other third-party payors, and there is not enough information available to derive out-of-pocket costs alone. The overall PCE estimates for health services, including insur- ance and related payments, are several billions of dollars higher than the corresponding SSA levels for the 1972-73 period, and nearly all of the difference is in the hospital care category. However, the PCE estimate for hospital services includes nursing home care and medical research financed by private and philan- thropic sources, and these additional items could account for a good part of the difference. Although there is some uncertainty about the hospital figure, there does not otherwise appear to be any basic inconsistency between the PCE and SSA estimates for direct medical care. Comparisons of Expenditure Categories Because of classification and conceptual differences, it is not possible to study the relationships between the survey and independent estimates in any appreciable detail. A few broad comparisons can be made, however, and these are presented in table 10-3. For this purpose, the "best" survey estimate is used. Where there was no significant statistical difference between the quarterly panel and diary levels, the former was chosen as the "best" estimate because it had the smaller coefficient of varia- tion. For categories in which there was a significant difference, the estimate that was generally closer to the independent levels was selected. Discounting many of the HIS levels as probably overstated, there appears to be a reasonably close correspondence between the survey and independent estimates for most categories. 6 The exception is for hospital services, for which the survey estimate seems to be substantially understated. It is possible that this disparity is exaggerated because of classification problems; some services in the survey were assigned differently from those in the independent sources. However, it is also possible that the more pervasive influence of insurance and other third-party payors in the hospital sector made it more difficult tc obtain complete and accurate reporting through the survey mechanism. Unlike the other categories, only one independent figure— the HIS estimate— is available for purposes of appraising the survey estimate for health insurance premiums. Since such premiums 5 See, for example, Marquis et al., op cit. ^The similarity for drug items, both prescription and over-the-counter drugs, was previously noted in chapter 4. are usually paid on a more-or-less regular basis (often through payroll deductions), it is less likely that the HIS level would be exaggerated, as appears to be the case for most other estimates from that source. The ratio of the survey to the HIS estimate for health insurance premiums is actually slightly in excess of unity, but— considering the problems of measurement— the best judgment is that the two sources are not substantively in dis- agreement. Timing Differences In previous chapters there has been considerable discussion of the possible effect of the time reference, or recall period, on the accuracy with which expenditures are reported. In most cases, it was concluded that a reduction in the recall period, if practi- cable, might achieve some improvement in reporting. The time reference used for health expenditures, as mentioned earlier, was the 6-month period preceding the month of inter- view. One of the more evident alternatives that might be ex- plored would be to reduce the recall period to 3 months. To examine this possibility, estimates were developed based only on expenditures reported in the final 3 months of each 6-month period. Unfortunately, the timing of the interviews was such that it was not possible to construct such estimates so as to be unbiased from a seasonal standpoint, in fact, as previously noted for home furnishings (chapter 7), the alternate estimates are biased upward because the month of December is overrepresented in "second-quarter" composites. Apparently, the level of health expenditures reported for December is well above the average for other months, probably because many large bills are paid at that time for tax deduction purposes. In addition to seasonal biases, the alternate "second-quarter" estimates might be exaggerated because of telescoping— the tendency to report an expenditure as having occurred in the more recent part of a recall period although it actually took place earlier. Although overstatements resulting from telescop- ing are often offset by understatements arising from a tendency to omit certain expenditures due to memory loss, the net balance may still be biased upward for expenditures of the type covered in this report. A comparison between the original and alternate estimates is presented in table 10-4 for major categories of health expendi- tures. Although the alternate estimates are higher in each instance, as might be expected, the implications of the differ- ences are not entirely clear. Even the seasonal and telescoping biases do not exaggerate the survey results for hospital services enough to make up for the gap between survey and alternate estimates. For the other categories, the evidence is that a reduc- tion in the time reference, without the usual "bounding" controls to overcome telescoping, would probably result in an overstatement of expenditures. What is indeterminate, however, is whether, after instituting such controls, there would be suf- ficient gains to warrant the effort and cost of collecting these expenditures on a quarterly basis. 42 On balance, in spite of the multitude of problems associated with measuring health expenditures, it appears that the 1972-73 survey was about as successful in this respect as most previous undertakings in this field. Of course, it is possible- that the difficulties are so great that none of these efforts, including the present one, is fully adequate. The one sector where the present survey is possibly less effective than other methods is the heavily insured hospital-service category. Table 10-1.— Annual Aggregate Out-of-Pocket Health Expenditures, by Major Category: Quarterly Panel and Diary Operation, 1972-73 Category Quarterly panel' Amount Estimated (millions coefficient of of dollars) variation (percent) 2 $2,713 5.0 8,896 2.0 5,284 2.0 1,913 4.0 3,699 2.0 11,811 2.0 Diary operation 1 Amount Estimated (millions coefficient of of dollars) variation (percent) 2 $1,900 15 9,183 5 5,318 5 2,032 15 5,019 5 2,880 (NA) Hospital services 3 Physician, nursing and medical services 4 . . . . Dental care Eye care 5 Prescription drugs • Health insurance premiums, including medicare (NA) Not available. 1 Estimates based on reweighted original survey data tapes and do not incorporate editing changes which may have been made at later stages of proces- sing by the Bureau of Labor Statistics. The survey estimates in this report may, therefore, differ somewhat from those already or to be published by that agency. Quarterly panel data are based on the average for calendar years 1972 and 1973, diary estimates on the average for fiscal years, July 1972-June 1974. 2 Preliminary. 3 Includes room and board, x-rays and laboratory tests in hospital, professional services included as part of regular hospital charges, etc. 4 Includes physician services in hospital where billed separately; physician services outside of hospital (except eye examinations); private nursing ser- vices; chiropractor services; other medical services outside of hospital such as x-rays, laboratory tests, physical therapy, etc. 5 Includes eye examinations and dispensing of eye glasses and contact lenses. Table 10-2.— Annual Aggregate Out-of-Pocket Health Expenditures, Excluding Cost of Drugs and Sundries or Health Insurance Premiums: Quarterly Panel and Diary Operation Compared to Independent Data Sources, 1972-73 Source Total expenditures' Amount (millions of dollars) Estimated coefficient of variation (percent) Estimated expenditures for nursing home care (millions of dollars) Total excluding nursing home care (millions of dollars) Survey estimates 2 Quarterly panel Diary operation Social Security Administration (SSA) estimates 3 Health Interview Survey (HIS) estimates (National Center for Health Statistics) 4 Center for Heelth Administration Studies (CHAS) estimates (University of Chicago) 5 $18,916 18,682 22,566 26,903 20,823 2.0 5.0 3.0 5.0 $ 110 250 3,332 $18,806 18,432 19,234 (NA) (NA) •Amount of nursing home care included in total expenditure estimates from these sources is probably quite small (of the order of magnitude shown for expenditure survey estimates), but no specific information is available on the amount. (NA) = Not available. 'includes hospital and nursing home care, physicians' services in and out of hospital, dental care, other professional services, and eye care and services. 2 Quarterly panel data based on average for calendar years 1972-73; diary estimates on average for fiscal years, July 1972-June 1974. Estimated co- efficients of variation are preliminary. 3 Based on aggregates prepared by that agency for fiscal year, July 1972-June 1973. Unlike the survey estimates, which relate to the civilian non- institutional population, the SSA estimates cover the entire population, which accounts for the relatively large amount shown for nursing home care. The estimates are not derived in a manner which permits calculation of variance measures. 4 Based on average of HIS survey data collected for calendar years 1970 and 1974. 5 Based on CHAS survey data for 1970 updated to 1972-73 by two methods: (1) percentage change for each major expense category as indicated in SSA estimates, and (2) percentage change for each major category as indicated in HIS estimates. CHAS estimates in table represent average of data derived from the two methods and essentially assume that the relationship between CHAS and the two other independent sources was the same in 1972-73 as it was in 1970 43 Table 10-3.— Annual Aggregate Out-of-Pocket Health Expenditures, by Major Category: "Best" Survey Estimates Compared to Independent Sources, 1972-73 Category 1 "Best" survey estimate 2 Amount (millions of dollars) Source (QP= quarterly panel; D=diary) Estimated coefficient of variation (percent) 4 Ratio of survey estimates to: : Social Security Administration (SSA) estimates Health Interview Survey (HIS) estimates Center for Health Administration Studies (CHAS) estimates Personal Consumption Expenditures (PCE) I GNP accounts] Hospital services Physician, nursing, eye care, and other professional services Dental care Drugs and sundries Prescribed drugs Health insurance premiums including medicare $2,713 QP 5.0 .752 .541 .777 (NA) 10,809 QP 2.0 1.049 .734 .963 (NA) 5,284 QP 2.0 .993 .738 .866 (NA) 8,076 D 5.0 1.034 (NA) 1.323 1.002 5,019 D 5.0 (NA) .862 (NA) 6 1.032 11,81 1 QP 2.0 (NA) 1.100 (NA) (NA) (NA) = Not available. See footnotes 3, 4, and 5 of table 10-1 for a description of the services included in the first two categories in this table. Note that eye care is combined with physician and other services in this table. 2 In cases where the quarterly panel and diary estimates were not significantly different, the former was chosen as the "best" estimate because it had the smaller sampling variance. Where the estimates were significantly different, the one generally closer to the independent sources was selected for this purpose. 3 See section of text entitled "Previous efforts to measure health expenditures" and footnotes 3, 4, and 5 of table 10-2 for a description of these inde- pendent sources. 4 Preliminary. Coefficients of variation would be slightly higher for the H IS estimates and perhaps 2 to 3 times higher for the CHAS estimates used in deriving the ratios in this table. SSA and PCE estimates are not derived in a manner which permits computation of variances. 5 CHAS figure used in this computation represents the smaller of two estimates derived from that survey. The higher estimate was exaggerated by the inclusion of one extreme sample report. The hospital services category in CHAS includes nursing home care, which was relatively small for the population covered. Based on survey and PCE data for 1972. Table 10-4.— Annual Aggregate Out-of-Pocket Health Expenditures for Hospital and Professional Services: Quarterly Panel Estimates Based on Data for All Months and on Reported Expenditures in Final 3 Months of Each 6-Month Period, Compared to Independent Data Sources, 1972-73 Quarterly panel Ratio of survey estimates to estimates (millions of dollars) Social Security Administration (SSA) estimates Center for Health Administration Studies Based on data for all months Based on data for reported expenditures in final 3 months of each 6-month period (2nd quarter estimates) 2 Category' Data for all months "2nd quarter" estimates (CHAS) estimates Data for all months "2nd quarter" estimates $2,713 $3,077 10,809 12,675 5,284 6 333 .752 .853 1.049 1.230 .993 1.190 .777 .881 Physician, nursing, eye care, and other professional services. . Dental care .963 1.129 .866 1 .038 'inclusions in these categories same as for table 10-3. 2 Data weighted to represent annual totals, essentially by doubling the weights for each reported expenditure. CHAPTER 11 Education, Travel, and Miscellaneous Expenses Previous chapters have covered expenditure categories ac- counting for more than 90 percent of all consumer outlays for products and some 75 percent of those for services. This chapter deals with a number of the remaining categories, including education, travel and transportation, and miscellaneous prod- ucts and services. Expenditures for Education An assessment of the survey results for education is handi- capped by the relative unavailability of comparable independent data and by the complications introduced by scholarships and other forms of student aid. The quarterly panel was the prin- cipal survey source for this information; families were asked semiannually about outlays during the previous 6 months for tuition, school housing and living costs, books and supplies, transportation, and other expenses, by type and level of school. Information was also obtained on the amounts, if any, reim- bursed by employers and on noncash scholarships or similar aid. Although various estimates may be derived from this composite of information, the principal survey data represent actual pay- ments made by the family, presumably including those made possible by cash scholarships, loans, or grants secured directly by the student or the family for educational purposes. The only mention of this subject on the diary record was the in- clusion of books and school supplies in one of the stipulated sections on the form; any other expenses were relegated to catchall sections. One limited comparative source that is available is the informa- tion on receipts from students by institutions of higher educa- tion prepared by the National Center for Education Statistics (NCES) of the Department of Health, Education, and Welfare. That agency currently does not prepare similar estimates for other levels of education. The Personal Consumption Expendi- tures (PCE) estimates for education prepared in conjunction with the GNP accounts are not usable for this analysis, because they represent the expenditures of the educational institutions themselves, including not only funds received from students but also receipts from endowments, gifts, and other sources. Also, the PCE data are confined to private schools and colleges. Table 11-1 compares the survey and the NCES data for student tuition and fees. It is evident, first of all, that the diary-based estimate is deficient in this respect, confirming the difficulty of measuring certain infrequent, high-cost items by record keeping. Although there may be some conceptual differences that affect the comparison, it appears that the quarterly panel estimates for higher education fall somewhat short of the NCES levels, a result not too surprising in view of the problems in covering college students in household surveys. Those students away at college and still supported by their families were considered as part of the family unit. Although the estimates of college expenses obtained from the family in these cases might be reasonably accurate, at least for tuition payments, there could be some inaccuracy if the student con- tributed part of the cost. Those students who were more-or-less independent were to be enumerated directly at their college quarters. However, the semiannual inquiries on educational expenses happened to be scheduled during the 3 summer months and the 3 winter months, when many students, if not most, might be residing elsewhere. Although missing informa- tion was to be picked up at the intervening quarterly interviews, there is no assurance that such information could be gathered in all cases for this population group. Although precise comparisons are not possible, the problems were probably considerably greater for other education-related expenses than for tuition. The quarterly panel estimate for housing and living costs for students away at college, for exam- ple, amounted to a little more than $2 billion annually. This was just about equal to the NCES figure for college receipts from students for housing and food, presumably in facilities operated by the institution. The survey figure should have been much higher, however, because it covered expenses in off-campus housing and eating facilities and also other living costs, such as laundry, local transport, entertainment, and the like. In spite of these difficulties, it appears, on balance, that an interview procedure such as that used in the quarterly panel may still be the best, if not only, option for estimating educa- tional expenses. Although the procedure would not be without problems, it might be better to interview students away at college at their school quarters instead of obtaining proxy information from their families at home. 1 An important con- comitant change would be to schedule such interviews during periods when there is maximum likelihood of finding students at their school quarters. Although the discussion thus far has been limited to the college group, these account for something like 75 to 80 percent of specific consumer outlays for education. Expenses for tuition and living costs for other levels of education would also appear to be best obtained from an interview process. However, for all 1 This is the procedure followed for college students in the Decennial Census. 44 45 cases, including college students, the diary procedure would probably be more effective for smaller expenses, such as books, school supplies, meals outside of school, boarding arrangements, and the like. 2 Travel and Transportation Although there have been a number of special surveys on vaca- tions and travel, the 1972-73 expenditure survey represents about the only comprehensive effort to date in the U.S. to measure individual family outlays for this purpose. The proce- dure used took advantage of previous experience in identifying the number of trips made for various reasons. In the quarterly panel, questions were asked each quarter about any trips lasting overnight or longer taken by any family members. Questions were asked about purposes of the trips, destinations, durations, means of transportation, and expenditures in considerable detail for various items. The "bounding" procedure described in previous chapters was used throughout to avoid telescoping (that is, duplicate reporting of the same trips in successive interviews). The survey estimates represent expenditures for overnight trips that ended during the survey year in question; they exclude trips taken entirely for business purposes or to or from schools and colleges. No direct comparisons are possible for travel expenditures. An indirect comparison— based on number of trips taken— can be made with the results of the National Travel Survey (NTS) taken in conjunction with the 1972 Census of Transportation. Since the accuracy of expenditure reporting often depends more on whether a particular expenditure is reported than on the precise amount of the expenditure, this comparison should be useful as an indicator. For purposes of the NTS, a trip was defined as travel involving a one-way distance of at 'east 100 miles, whether it lasted overnight or not. Business trips were included but not those to and from school. It is possible to adjust the NTS data to exclude trips not involving overnight stays and those primarily for business purposes. One remaining incomparability is that the expenditure survey estimates include overnight trips less than 100 miles from home and some trips primarily, but not exclusively, for business purposes. On the other hand, a series of identical trips (such as a person's visit to his or her family each weekend) was combined into a single trip in the expenditure survey; each would be considered a separate trip in the NTS. Table 11-2 compares the quarterly panel and the NTS data on number of trips in a few categories. Many more comparisons could be made, but the necessary detail is not yet available for the expenditure survey. The estimates are not significantly dif- ferent for any of the comparisons shown, although the expendi- ture survey data probably should be somewhat higher at least for the shorter trips. Nevertheless, in terms of trips at least, the data would not likely be very much out of agreement,, whatever reasonable assumption is made for the remaining incomparabilities. In a related sector, expenditures for recreational lessons (music, dancing, swimming, etc.) amounted to about $750 million annually in both the quarterly panel and diary operation and probably could be accommodated adequately in the latter. One limited aspect of travel-related costs for which another comparison is possible is housing expenditures. The quarterly panel estimate of trip-related expenditures on hotels and other living quarters amounts to $3.3 billion annually in 1972-73. A Personal Consumption Expenditures (PCE) estimate for hotels, motels, and trailer parks— which is not necessarily pre- cisely comparable with the expenditure survey figure— was not significantly different ($3.1 billion annually). In a related sector, table 1 1-3 compares the survey and the PCE estimates for various categories of public transportation. The diary estimates are used in this instance because they are the only comprehensive survey measures for this sector. The com- parison is somewhat obscured by the presence in the diary estimates of a sizable residual group, attributable to many in- complete or inadequately described entries in the diary record. Even making reasonable allowance for this residue, it is evident that the survey estimates are considerably understated relative to PCE for local transport but less so for other kinds of trans- portation, especially airline fares. These relationships are con- sistent with previous findings: expenditures, such as local trans- port, likely to be made by individual family members might be incompletely known and sometimes overlooked by the home- maker, usually the recordkeeper for the family. Expenditures for airline and rail travel, on the other hand, would generally represent more significant events, less likely to escape the attention of the recordkeeper. Miscellaneous Products and Services Table 11-4 compares survey and independent estimates for a group of miscellaneous products and services. The selection has no significance other than representing items for which reasonably valid comparisons appeared to be feasible at this writing. A more extensive and detailed analysis should be possible after the present PCE benchmark adjustment (to 1972 Census data) is completed. Most of the categories shown represent those in which expendi- tures are relatively large in terms of unit costs. The results,, not surprisingly, are consistent with previous findings for items of a similar nature. In the majority of instances, the quarterly panel provides either the only or the "best" survey data, and the estimates from the panel correspond reasonably well with the independent PCE levels. One exception to this correspondence is bicycle purchases. In this case, the apparent survey underestimate may be attributable in part to the fact that the generally successful inventory method was not used for this item. Also, the difference may be exaggerated because of greater-than-usual uncertainties about the validity of the PCE estimate in a sector increasingly dominated by imports. The marked understatement for watches and jewelry 3 could arise from a variety of causes, among them the irregularity of many of these purchases and the fact that so large a proportion may represent gifts. It is also possible that jewelry expenditures could be deliberately concealed because of reluctance to reveal the appearance of extravagance or even because of concern about the security of expensive possessions. These expenditures were collected on a quarterly basis in the interview panel. 46 ■ Table 11-1.— Annual Aggregate Expenditures for School Tuition and Fees: Quarterly Panel and Diary Estimates Compared to National Center for Education Statistics Data, 1972-73 1 Source All schools Amount (millions of dollars) Estimated coefficient of variation (percent) 3 Colleges and universities Amount (millions of dollars) Estimated coefficient of variation (percent) 3 Survey estimates 5 Quarterly panel Diary operation National Center for Education Statistics (NCES) data (U.S. Dept. of Health, Education and Welfare) 4 $7,243.6 3,171.0 (NA) 3.0 9.0 $4,977.3 (NA) 6,035.0 3.0 (NA) = Not available. 'Quarterly panel estimates are averages for calendar years 1972 and 1973. Diary estimates are averages for 2 fiscal years, July 1972-June 1974, and NCES estimates are averages for 3 fiscal years, July 1971-June 1974. 2 Survey data are derived from original data tapes and do not incorporate editing changes which may have been made at later stages of processing by the Bureau of Labor Statistics. The estimates may differ somewhat, therefore, from those already published or to be published by that agency. Preliminary. 4 See appendix A for description. Table 11-2.— Annual Aggregate Number of Nonbusiness Overnight Trips: Quarterly Panel Compared to National Travel Survey, 1972 Type of trip Quarterly panel 1 Number of trips (millions) Estimated coefficient of variation (percent) 3 National Travel Survey (1972 Census of Transportation) 2 Number of trips (millions) Estimated coefficient of variation (percent) 3 Total trips Trips within U.S. Less than 500 miles round-trip 500 or more miles round-trip . Foreign trips Destination not reported 137.9 98.0 31.9 7.3 0.8 4.5 4.5 5.5 7.0 135.7 94.9 32.7 8.2 3.0 3.0 4.0 5.0 1 See footnote 2, table 11-1, for qualifications concerning survey data. 2 See appendix A for description. 3 Preliminary. Table 11-3.— Annual Aggregate Expenditures for Public Transportation: Diary Estimates Compared to Personal Consumption Expenditures, 1972-73 Diary operation 1 Personal Consumption Expenditures (PCE) (millions of dollars) 3 Category Amount (millions of dollars) Estimated coefficient of variation (percent) 2 Total $4,888.6 1,042.9 457.4 256.8 2,360.8 770.7 6.0 8.0 10.0 15.0 8.0 $6,318.5 2,135.0 838 5 368.5 2,812.5 164.0 1 See footnote 2, table 11-1 for qualifications concerning survey data. Diary estimates are averages for fiscal years, July 1972-June 1974. 2 Rough preliminary estimates. 3 See appendix A for description. 4 Includes intercity as well as local bus. 5 Includes commuting as well as intercity travel. 4 7 Table 11-4.— Annual Aggregate Expenditures for Miscellaneous Products and Services: Quarterly Panel and Diary Operation Compared to Personal Consumption Expenditures Estimates, 1972-73 Quarterly panel 1 Diary operation 1 Ratio of survey Category Amount (millions of dollars) Estimated coefficient of variation (percent) 2 Amount (millions of dollars) Estimated coefficient of variation (percent) 2 estimates to Personal Consumption Expenditures (PCE) 3 Quarterly panel Diary operations Products $179.5 20.0 261.2 9.0 637.2 4.0 (NA) (NA) (NA) .945 Other musical instruments, new (excluding organs)* . . Bicycles' .899 .651 Services Furniture repair and Electrical appliance repair 5 2,815.7 726.5 2,411.2 589.1 2,038.1 3.0 7.0 5.0 7.0 5.0 $2,340.9 255.2 1,081.4 669.4 1,627.5 6.0 * • * * * <- .557 .883 .939 1.102 .802 .463 .310 .421 1.252 .640 (NA) = Not available. •Comparison for 1972 only; others based on 1972-73 average. "Not available, but probably 2 to 3 times the size of the comparable quarterly panel coefficient. 1 See footnote 2, table 11-1 for qualifications concerning survey data. 2 Preliminary. 3 See appendix A for discussion of PCE data. 4 Includes cemetery plots as well as funeral parlor and burial expenses. 5 Includes radio, television, and other sound equipment repairs as well as those for other household electrical appliances. Survey figures include cost of service contracts for these purposes as well as direct expenditures. CHAPTER 12 Conclusions and Recommendations In any complex new endeavor— and survey undertakings are certainly no exception— it would be unrealistic to expect all program objectives to be accomplished at the first trial. In fact, success at that early stage may be reasonably measured by the degree to which at least tolerable results have been achieved. Using that yardstick, the 1972-73 survey can probably be judged a success. Perhaps even more important than the im- mediate result is the question of whether enough has been learned from the experience to help shape a significantly im- proved product for the future. Although numerous uncertainties remain, a good many promising indications have emerged from this preliminary evaluation. General Conclusions for Diary Operation Many of the findings from the 1972-73 survey were predictable on the basis of previous experience. For example, larger expen- ditures and recurrent outlays were better reported in the quar- terly panel. Expenditures made primarily by the homemaker or on a frequent basis were more successfully reported in diary form. An unexpected outcome was the relatively effective per- formance of the diary procedure in some sectors, such as health and housing, in which the quarterly panel had been assumed to be the only realistic option. At the same time, many deficiencies were observed in the diary results, even in categories in which the diary procedure had been deemed to be the more appro- priate source. The specific findings of this report with respect to the diary procedure led to several general conclusions and recommenda- tions: 1. Use diary procedure as primary source unless evidence clearly dictates otherwise.— A general rule of thumb sug- gested by the findings is that unless a clear-cut reason exists for using an interview procedure, such as the quarterly panel, it is better to depend on the diary ap- proach for a given expenditure category. The diary would be a dubious source for items with exceptionally large variances (vehicles, appliances, furniture, etc.) or where unusual payment arrangements might require special questioning (insurance paid through payroll deductions, mortgage payments made automatically through bank accounts, hospital bills paid largely, but belatedly, by insurance, etc.). The diary might also be inferior to an interview process in cases where a composite set of ques- tions may be necessary to obtain complete responses, as for example, in cataloging trips and vacations. In most other instances, the diary procedure appears to be at least as good a risk as the interview approach, and it is probably a less costly one. Clearly, however, a number of modifica- tions and improvements in the diary procedure are needed in order to overcome some rather evident deficiencies. Limit the range of items any one family would be asked to report.— Using a diary covering all items of expenditure, as was done in the 1972-73 survey, may have certain theoretical benefits, but this approach considerably limits the ability to improve the overall process enough to satis- fy all survey requirements. Evidently, one of the reasons for more success in covering food purchases than in cover- ing other small items in the 1972-73 diary was the much greater amount of space and attention given to food. Moreover, for all categories, including food, there was a considerable undifferentiated residual group, resulting mainly from incomplete or inadequate entries which could not be classified in detail. These residuals detracted materially from the usefulness of the results. The general lack of space and the inability of interviewers to focus on so wide a range of items in reviewing the diaries probably largely accounted for this deficiency. It would be impracticable to provide adequate space and annotation for all categories on a single form, and one rather evident solution would be to limit the range of items any one family would be asked to report. For example, one subsample might be asked to report only on food and other supermarket products, a second on cloth- ing and household linens, a third on health-related ex- penditures, and so forth. There is obviously some practical limit to the number of subsamples that could be operated simultaneously. A good deal of thought and some ex- perimentation would be needed to devise a workable plan. Probably even more important than space considerations is that the use of this kind of specialized approach would make it feasible to provide a more focused set of check questions and procedures at the time of diary pickup. This would overcome some of the disparities noted in the present survey (such as underreporting of certain food items relative to others). Modifications of the present procedure might include more use of a combination of interviewing and recordkeeping. A good deal of develop- mental work would be required to devise effective check- ing procedures. Vary length of recordkeeping periods.— The 1972-73 survey called for each sample family to keep records covering all items of expenditure for two weeks. If special- ized subsamples are developed as proposed above, it is 48 49 obvious that either a larger overall sample would be need- ed or much higher sampling variances would have to be accepted. One way out of this dilemma would be to vary the length of the recordkeeping period, depending on the variances of the subjects covered for a given subsample. For example, for a low-variance category, such as food, it might even be possible to reduce the recordkeeping period to 1 week or 10 days. For most categories, such as clothing expenditures or health costs, however, an in- crease in the period of recordkeeping up to 3 months or longer might be considered. If only a limited set of items is covered, the reporting burden might be sufficiently reduced so that cooperation might be extended. In the system suggested, less costly collection methods, such as having respondents mail in completed diaries on a periodic basis (monthly, semimonthly, etc.) would be more practi- cable. Of course, provision would have to be made to follow up (by mail or telephone, if possible) nonrespon- dents or to carry out special checking for returns that did not meet prescribed standards. Provide separate diaries, where indicated, for individual members.— In most cases, one person, the homemaker, probably maintained the diary for the entire family in the 1972-73 survey. Not surprisingly, the results clearly were better for the kinds of expenditures made by the home- maker than for those likely to be made by other family members. One possible way of obtaining more consistent results, if the expenditures to be reported are dispersed, would be to provide all family members over a certain age (perhaps 12 and over) with separate diaries in which to record their individual disbursements. 1 For this pur- pose, the diaries could be briefer and less forma! than the main family record. Experimentation with various ver- sions would obviously be important in developing a procedure of this kind. Reconsider the matter of providing monetary or other incentives for cooperation.- An experiment was con- ducted in the early stages of the 1972-73 survey on whether an offer of cash payments would materially im- prove cooperation in maintaining diaries. The results were inconclusive, and the incentives were dropped from the procedure. However, most previous experience sup- ports the notion that both cooperation and adequate reporting are benefitted by inducements of this kind. If greater dependence is to be placed on diaries in a continuing operation, as is being proposed, it would seem especially important to reconsider offering incentives and to experiment with alternative approaches. The rewards should, of course, be attuned as closely as pos- sible to achieving the main objectives. !f, for example, it is important for individual family members to keep separate records, the incentives should be offered only if all such members agree to cooperate. Such cooperation would This procedure has been used successively for some years in the continuous consumer expenditures survey in the United Kingdom. presumably promote intrafamily pressure for individual compliance. Similarly, if cooperation over extended periods is requested, a useful approach is to offer a small reward for each subperiod with a sizable bonus for com- pleting the entire cycle. Special consideration for prompt and complete returns by mail (if used) might be still another element. Payment of incentives might also be made contingent on respondents retaining cash-register tapes, bills, and other evidence of expenditures which could be consulted in reviewing the diaries. The incentives, of course, do not necessarily have to be monetary but could include gifts of various kinds, publications, or chances for prizes. 6. Continue exploration of timing biases.— Perhaps the most conclusive survey finding was that the traditional bias in diary operations was affirmed; that is, a higher level of expenditures is reported in the earlier as opposed to the latter stages of the recordkeeping period. In the present instance, the first-week expenditure estimates exceeded those for the second week in almost every category. More detailed information on this subject, including data for separate days of the reporting period and for various al- ternative combinations of days, will be provided in a later report. Although these details will hopefully provide further insights into the reasons for these differences, it seems likely that further research will be necessary to assess the implications of this phenomenon from the standpoint of deciding on recordkeeping periods and deriving estimates from diaries. General Conclusions for Quarterly Panel Although the deficiencies in the quarterly panel results could be attributed to a variety of causes, the fatigue caused by the voluminous questionnaire was undoubtedly a prime factor. Perhaps the single most important step that could be taken to improve the results would be a significant reduction in the scope and content of the inquiry. Extending the use of the diary pro- cedure would accomplish a good deal in this direction. Such a change, however, would impair one of the auxiliary objectives of the survey— the ability to analyze the consumption patterns of individual families across a wide range of products and services. From the standpoint of the operation as a whole, the main general conclusions with respect to the quarterly panel follow: 1. Restrict coverage to essential items.— Mention has already been made of the kinds of Items that should p'obably be retained in the interview procedure— high variance items, those requiring special questioning, and perhaps certain topics for which there were efficient composite inter- viewing packages. On this basis, the quarterly panel would probably cover not only vehicles, appliances, and furni- ture but also larger clothing items (e.g., coats and suits), other major furnishings (floor coverings and drapes), hospital services, school and college tuition, mortgage payments, moving and funeral expenses, insurance, sub- scriptions and membership fees, travel and vacation costs, and the like. Obviously, a more detailed review and analysis would be necessary to select the optimum set. Extend "inventory" approach.— The effective use of the so-called "inventory" approach for appliances suggests that it be extended to other categories. Examples would be major furniture items, floor coverings and drapes, and certain kinds of durable equipment previously excluded. The evidence also suggests that, in the interest of reducing sampling variances, consideration be given to pooling the two estimates obtainable for a given year (or other time period) under this approach. Retain flexibility in reference periods.— The procedure used in the 1972-73 survey drew on the well-documented principle that the larger the expenditure, the longer it can be recalled reliably. Thus, inquiries were made for specific categories in the interview panel on either an annual, semiannual, or quarterly basis, depending mainly on expenditure size. For the most part, the range of those choices was confirmed by the results. In a few cases, it appeared that a reduction in the recall period from 6 months to 3 months might be beneficial for some of the smaller items (certain household furnishings, a few of the smaller appliance groups, etc.). However, there did not appear to be any need to curtail the time reference for the larger expenditures. Longer recall periods for these items would serve to minimize the number of occasions any given family would be asked about the same pur- chases. 2 Retain analytical opportunities, where possible, with respect to individual family behavior.— The proposed re- duction in the content of the quarterly panel interviews and the possible introduction of specialized diaries means that only very partial expenditure data would be available for individual families, either for annual or for shorter periods. It would be desirable, if this could be clone with- out jeopardizing the basic expenditure data, to retain as much of the ability as possible to study individual con- sumer behavior. Some of the largest items, which distinguish one family's expenditures from another's, would be retained on the quarterly panel under the proposal, but most categories would not be covered. A first step to close the gap might be to add a few questions, perhaps on an annual basis, to identify exceptional expenditures in other sectors, such as large orthodontic bills, major house additions, purchase of expensive paintings, and the like. Since appliances, furnishings, and similar items are purchased very infrequently, repetition of the same questions on a frequent basis (e.g., quarterly), where it is not clearly necessary, is likely to create a consider- able annoyance (and waste of survey capacity) without achieving any productive results. Beyond that point, a further, but admittedly experimental, step would be to obtain certain limited information in the quarterly panel on behavioral patterns. This step would help in imputing data from the diary operation or other sources to the individual family record. For example, in the health sector, questions could be asked on frequency of doctor or dental visits, consistent use of costly drugs, etc. Answers to such questions, together with demo- graphic and socioeconomic characteristics, could provide a basis for imputing from the diary or another source a value for health expenditures for the family. The topics covered could vary from one quarter to another, so that these behavioral questions would not add excessively to the quarterly panel interviews. Obviously, considerable research and testing would be needed before embarking on a large-scale endeavor of this kind. Possible Use of Other Data Sources When independent estimates were collected for purposes of comparison with the survey results, it became apparent that there are a variety of such sources that should be taken into account in developing a continuing data system in this field. For example, the Health Resources Administration of the U.S. Public Health Service is instituting a detailed national household survey on health expenditures. The Census Bureau operates a national annual housing survey containing some housing expenditure data and another survey on expenditures for home repairs and alterations. That agency is also launching a household survey on trips and vacations. There are also many market research and other private operations of this kind, although of rather uneven quality. Some of these operations are or have been on a continuing basis, but the status of others is uncertain. The main point is that, in a system such as that suggested in this report, there is no reason that data from other sources of this kind could not be substituted for those from a continuing comprehensive expenditure survey, provided a set of matching demographic and socioeconomic characteristics is available. The advantage of specialized surveys, assuming they are of adequate size, con- tinuity, and technical quality, Is that particular subjects general- ly can be explored in considerably greater detail, using method- ologies especially designed for that purpose. Need for Methodological Research Several instances have already been cited whereby further methodological research would be essential in shaping a more effective data system for consumer expenditures. Although the need*, are varied, it appears that the highest priority would be to experiment with modifications in the diary procedure, such as testing the use of specialized diaries and diaries for individual family members, developing special checking procedures to overcome reporting deficiencies, exploring the feasibility of 51 varying lengths of diarykeeping periods, considering less expen- sive collection methods, and reexamining the use of monetary or other incentives for cooperation. Some longer range pos- sibilities should also be investigated. For example, respondents might be asked to record the "universal product codes" now appearing on most canned and packaged supermarket and drug store items (likely to extend to many others) in the interest of achieving more precise and consistent product classification. In a different vein, when collecting diaries, interviewers might seek permission to take brief shelf inventories in an effort to locate the kinds of products that tend to be overlooked inad- vertently by respondents. In more sensitive areas, such as alcoholic beverages, tobacco products, jewelry acquisitions, and the like, experimentation with techniques used in other sensitive fields would be in order. Finally, some further experimentation should be carried out on timing biases (e.g., first-second week differences), which are characteristic of such surveys. Since the survey findings support retaining an interviewing capability as a companion to the diary procedure, some further methodolocial research relating to the quarterly panel would also be in order. One fruitful area might be the extension to additional sectors of the inventory approach used for appli- ances in the 1972-73 survey. Another might be exploring less costly collection procedures in this instance, such as use of mail questionnaires or telephone interviews for certain of the quarterly contacts following the first visit. Mention has already been made of the need for research in the event that considera- tion is given to imputing expenditure data from various sources to individual family records. APPENDIX A Description of Independent Sources Various independent sources of data have been used for com- parison with the survey results. A description and discussion of these follow: Personal Consumption Expenditures (PCE) Com- ponent of the Gross National Product (GNP) Accounts The principal source of independent estimates used in this re- port is the Personal Consumption Expenditures (PCE) com- ponent of the Gross National Product (GNP) accounts. These estimates, in effect, represent the market value of goods and services purchased by persons and nonprofit institutions in the United States. A description of the methods used in deriving the PCE estimates occupies one hundred or more pages of text, 1 and only a brief summation will be presented here. The principal starting point is data on the production value of goods from the quinquennial census of manufactures and other sources and data on the value of services from a variety of sources. Through input-output analysis, the flow of commodities and services is traced through various channels in the economy; and to arrive at the final market value, cost and profit margins are added to the production value at each stage. Numerous assumptions are needed to achieve these complex transformations, and there is no way of assessing the accuracy of the processes. For many of the comparisons made, especially those of a sum- mary nature, PCE estimates are prepared by the U.S. Depart- ment of Commerce on a current basis. Those used in this report, however, have not yet been adjusted to the most recent bench- mark (1972 Census data), a step which is currently nearing completion. For more detailed categories, PCE estimates are ordinarily de- veloped only for benchmark years, with the most recent, available at this writing, relating to 1967. To permit more detailed comparisons, the author performed an updating of the 1967 PCE estimates using a rather straightforward approach. The procedure was to project the 1967 PCE figures to 1972 by ap- plying the percentage change in production value for the corre- sponding items as shown in the 1967 and 1972 Censuses of Manufactures. A similar procedure was followed for imports (shown separately in the detailed PCE estimates), in this case using census foreign trade statistics for the two dates. The main assumption in this method is that the marketing margins (per- centage difference between production and market value), as calculated in PCE, did not change between 1967 and 1972, U.S. Department of Commerce, Office of Business Economics (now Bureau of Economic Analysis), "Development of National income Measures," Supplement to Survey of Current Business, Washington, 1954. probably a conservative approach, since marketing and related costs have reportedly risen especially rapidly. Of the numerous conceptual differences between the survey and PCE estimates, one is the inclusion in PCE of expenditures by nonprofit institutions (not a major factor in most comparisons). A discussion of relevant conceptual differences for various expenditure classes is included in the appropriate detailed chapters in this report. 1972 Census of Business Merchandise Line Data (Cens. Bus.) In the 1972 Census of Business, respondents were requested to provide a distribution of retail sales in their establishments in a number of broad merchandise lines. 2 For establishments likely to carry a particular line in any quantity, additional detail was often requested. Although response rates are lower than most others for this aspect of the Census, compliance amounted to 75 percent or more in most cases where these data were used for comparative purposes. In compiling merchandise line estimates, the Census Bureau adjusts the data for nonreporting. The census retail trade data include sales to businesses and other nonhousehold users, although these are not believed to be large for those categories for which the data were used for comparative purposes in this report. Conceptual or coverage differences for particular expenditure classes are discussed, where applicable, in the appropriate detailed chapters. 1972 Census of Selected Service Industries (Cens. Serv.) As part of the 1972 economic censuses, a census was taken of selected service industries. For such industries, data on gross receipts, along with many other items, are available. 3 For those industries catering almost exclusively to the private population (and where no significant part of the receipts can ordinarily be attributed to business expenses, e.g., barber and beauty shops, motion picture theaters, etc.), the data on receipts were used directly in comparisons with the survey estimates. For more complex industries (e.g., laundry and dry cleaning), more detailed information on the distribution of receipts by types of "U.S. Department of Commerce, Bureau of the Census, 1972 Census of Retail Trade, Merchandise Line Sales, Report No. RC 72-L, Wash- ington, September 1975. 3 U.S. Department of Commerce, Bureau of the Census, 1972 Census of Selected Service Industries, Subject Series, Report No. SC72-58, Washington, December 1975. 53 54 services is available for those establishments with paid employ- ees. It was possible to derive estimates from this detailed infor- mation which were reasonably comparable to the survey data. For this purpose, however, it was necessary for the author to make a small adjustment in the original census statistics to ac- count for receipts of establishments without paid employees. Economic Research Service (ERS), Department of Agriculture Estimates Annual estimates of expenditures for farm-produced foods used by the U.S. civilian population are prepared by the Economic Research Service (ERS) of the U.S. Department of Agriculture for 15 major subcategories of food products. 4 These estimates are developed, basically, from production estimates compiled by the Department and allowing for imports and exports and noncivilian use. Expenditure estimates are derived from the net production data by applying average retail price data from the BLS Consumer Price Index, for commodities covered in that Index, and price data from the Department's Agricultural Research Service for other commodities. The ERS estimates provide a split between expenditures for home use and those in outside eating places for 7 broad categories (combinations of the 15 categories for which annual expenditure data are com- puted). The author extended this allocation to some of the additional categories by using the relationships shown in PCE at-home/away -from-home consumption. Panel Data from the Market Research Corporation of America (MRCA) A continuing national consumer panel of about 7,000 house- holds is canvassed by mail by the Market Research Corporation of America, on a weekly basis for a variety of food products and on a monthly basis for clothing, household textiles, and various other items. For the monthly items, in particular, a considerable understatement in reporting has been observed over the years. However, MRCA has developed adjustment factors in con- junction with its clients (based on comparisons of panel and sales data for specific categories) in an effort to overcome this deficiency. The MRCA data used here incorporate the effects of these adjustment factors, the accuracy of which is unknown. Social Security Administration (SSA) The Social Security Administration (SSA) prepares annual aggregates of national health expenditures in various broad categories— hospital care, physicians' services, drugs and drug sundries, etc. 5 Data are developed on the total cost of such services and products and on the source of funds for payment, whether directly by the consumer, or indirectly by health in- surance, government, philanthropic or industrial enterprises, etc. A variety of primary data sources are used in the prepara- tion of these estimates, including information on health pro- viders from Internal Revenue Service records and statistics compiled by the American Hospital Association and other private health-related organizations. As is customary for con- structed series of this kind, numerous transformations are made of the original data, and there is no way of assessing the validity of the results. Health Interview Survey (HIS), National Center for Health Statistics, U.S. Public Health Service On occasion (most recently for 1970 and 1974) an inquiry on health expenditures has been appended to the continuing Health Interview Survey (HIS) conducted by the Census Bureau under the sponsorship of the National Center for Health Statistics of the U.S. Public Health Service. 6 The procedure in these recent efforts has been, at the conclusion of the regular HIS interview, to drop off questionnaires on this subject which are to be self- administered and mailed in by respondents. A separate form is provided for each family member requesting estimated total expenditures in the previous year for a few broad categories- dental bills, doctors' bills, hospital bills, prescription medicine, eyeglasses, and miscellaneous items. Information on health in- surance premiums is also obtained. For comparison with the expenditure survey data, an average was taken of the 1970 and 1974 HIS estimates both of which were based on national samples of about 10,000 households. Center for Health Administration Studies (CHAS), University of Chicago A series of intermittent household surveys (the most recent covering 1970) on health expenditures has been conducted by the Center for Health Administration Studies (CHAS) of the University of Chicago. 7 The procedure was to conduct an extremely detailed interview with the family to identify all episodes during the year which might entail substantial health expenditures, such as hospitalizations, serious accidents, preg- nancies, chronic conditions, etc., and to inquire further about treatment and costs for each episode. A final series of questions addressed residual, smaller expenditures for the same period. About 3,800 households were covered in the most recent study. The 1970 CHAS data were updated to 1972-73 by using trend data reported in Social Security and Health Interview Survey data for comparable expenditure classes (see above). Annual Housing Survey (AHS), Bureau of the Census Since 1973, the Bureau of the Census has conducted a national Annual Housing Survey under the sponsorship of the Depart- ment of Housing and Urban Development. 8 The data collected U.S. Department of Agriculture, Economic Research Service, Marketing and Transportation Situation, Washington, August 1975 U.S. Department of Health, Education, and Welfare, Social Security Administration, "Compendium of National Health Expenditures Data," DHEW Pub. No. (SSA) 76-11927, January 1976. 6 U.S. Department of Health, Education, and Welfare, Public Health Service, "Personal Out-of-Pocket Health Expenses, United States, 1970," DHEW Pub. No. (HRA) 74-1518, Washington, June 1974. 7 Ronald Andersen, Joanna Kravits, Odin W. Anderson, and Joan Dudley, "Expenditures for Personal Health Services, National Trends and Variations (1953-1970)," DHEW Pub. No. (HRA) 74-3105, Washington 1973. 8 U.S. Department of Commerce, Bureau of the Census, "Annual Housing Survey: 1973," Current Housing Reports, Series H-1 50-73, Washington, April 1976 55 consist primarily of detailed characteristics of housing units and their occupants. A limited number of expenditure items is also included on rent, mortgage payments, and other housing costs; and, in the early years, some questions were also asked on expenditures for automobiles and selected household appliances. The 1973 expenditure estimates relate roughly to the 12-month period from autumn of 1972 to autumn 1973. The 1973 sample consisted of some 53,000 interviewed units in 461 areas through- out the country. Survey of Residential Alterations and Repairs (SORAR), Bureau of the Census For a number of years, as part of its construction statistics program, the Bureau of the Census has conducted a quarterly sample household survey to collect data on expenditures for alterations and repairs by owners of residential properties. 9 The survey uses a very detailed approach for these purposes, includ- ing a "bounding" technique, described earlier in this report, to avoid telescoping. Households are usually interviewed for 6 consecutive quarters for this survey, with the first quarter used mainly for set-up purposes and to screen out inapplicable cases. The interview sample consists of an effective sample size of about 6,500 owner-occupied households each quarter in 235 sample areas. In addition, about 4,000 nonresident owners of residential properties are canvassed by mail for these purposes. National Center for Education Statistics (NCES), Department of Health, Education, and Welfare Among its many institutional surveys, the National Center for Education Statistics (NCES) conducts an annual "Higher Education General Information Survey" (HEGIS), covering virtually all public and private institutions of higher education (2-year colleges as well as 4-year institutions and universities). 10 A considerable array of information is collected, including U.S. Department of Commerce, Bureau of the Census, "Residential Alteration and Repair, 1972 Annual Report— Part 1, All Residential Properties," Construction Reports, Series C50-72A, Washington, Oc- tober 1973. U.S. Department of Health, Education, and Welfare, National Center for Education Statistics, "Financial Statistics of Institutions of Higher Education," Report No, NCES 76-121 , Washington 1976. financial data on receipts of funds by source and expenditures by purpose. The data relate to fiscal years (July-to-June); and, for comparison with the 1972-73 expenditure survey data, it was decided that an average of HEGIS data would be appro- priate for the three fiscal years, July 1971 -June 1972, July 1972 June 1973, and July 1973-June 1974. Response rates for the HEGIS survey are geneially quite high (around 95 percent of the institutions contacted). The data from HEGIS on receipts from students for tuition and fees were regarded as reasonably comparable to the consumer outlays for this purpose, reported in the expenditure survey. National Travel Survey (NTS), Bureau of the Census The National Travel Survey was an adjunct to the 1972 Census of Transportation. 11 The survey was conducted with a national probability sample of some 24,000 households in 449 different geographic areas of the country. The sample was subdivided into three systematic subsamples with each canvassed quarterly on a different cycle of months (e.g., the first in February, May, August, and November; the second in March, June, Sep- tember, and December, etc.). Information was collected on numbers of trips taken and on various details about each trip. Mail questionnaires were used with provision for mail and tele- phone followup to handle nonresponse or to clarify question- able data. Cooperation averaged 90 percent for eligible house- holds. A.M. Best and Co. Aggregate data on gross insurance premiums for private vehicles, compiled from industry sources by A.M. Best and Co., 12 have been used for comparison with the survey data. These are also the base data used by the Commerce Department in devel- oping its PCE estimates of net premiums (gross premiums less claims paid) for vehicle insurance. U.S. Department of Commerca, Bureau of the Census, National Travel Survey, "Travel During 1972," Report No. TC-72-N, Washing- ton, September 1973. A.M. Best and Co., "Best Aggregate and Averages— Property-Liabil- ity ," Annual publication. APPENDIX B Facsimile of Diary Pages for One Day of the Reporting Period FIRST DAY ENTER DAY ^^ Note: If you need additional space for any category, use pages 18 and 19. FOOD AND BEVERAGES OFFICE USE ONLY Item (Describe the item purchased, such as wfio/c milli, T-bone steak, dried apricots, all purpose flour, saltmes. ere.) Number of cans, bottles, packages, etc. Net weight 01 volume per unit (Examples: 8 01., 1 or., 5 lbs., etc.) Is this item - (Mark anly one) Total cost (Exclude soles fo*J Fiesh Frozen Canned Other Dollars 1 Cents -51 on Dairy and Bakery Products (Indicate i( milk is delivered! A 6 1 12 13 14 J 1 i 2 ,3 ,4 c 1 12 I 3 | 4 n 1 12 13 14 E 1 12 13 14 F 1 12 13 14 G 1 12 13 14 H t 12 13 14 J I '2 1 3 1 4 X 1 '2 13 14 (*»al, Fllh, Md Poultry (Indicate the cut of meal) L ' I 2 ' 3 1 4 J M 1 |2 i 3 ,4 N 1 |2 | 3 i 4 P t 12 13 14 Q 1 12 1 3 '4 R ' I 2 1 3 I* 5 1 |2 ] 3 • j « T 1 [2 [3 | 4 U 1 [2 [S ,4 ~51 029 Fruits ind V«{«tsbl»s (Indicate if dried) A 1 12 13 4 $ B 1 |2 i 3 i 4 C 1 12 13 14 D 1 12 >3 14 E 1 [2 '3 '4 F 1 J2 J3 ]4 C 1 ,2 ,3 ,4 H 1 12 3 4 J 1 12 13 14 K 1 12 13 14 Bovoragos L 1 |2 '3 U % M t J2 1 3 |4 N 1 [2 3 \t P 1 2 3 4 Q 1 |2 ,3 ,4 R 1 12 13 14 - 51 037 All OtlMr Foods A 1 '2 '3 '4 i B 1 12 13 4 C 1 12 13 14 1 12 13 14 E t 12 13 14 F G 1 12 13 14 1 12 13 14 H 1 12 13 14 J 1 12 13 '4 K t >2 13 14 L 1 '2 '3 '4 M t 12 13 14 N t 12 13 14 P 1 12 13 14 a 1 12 13 14 1 57 Paje 5 FIRST DAY n None npc MEALS AND SNACKS PURCHASED AT A RESTAURANT, CARRY-OUT, ETC. OFFICE USE ONLY Item (Include breakfasts, dinners, school lunches, snacks, and drinks purchased al a restaurant, bar, vending machine, etc) Was this purchased at a - (Mark one) Total cost (Include tips) Restau- 1 Cafe- | °"^" ln ! Vending | 0h ' ar '< J ,e " a ;car,y-ou<; machine ; i i I I i i I i ~ 31 047 Dollars i Cents A 1 12 13 14 IS 1 1 1 1 $ B 1 1 2 13 14 IS 1 I 1 1 C 1 12 13 '4 IS 1 1 1 1 D 1 12 13 14 15 1,11 ALL OTHER PURCHASES AND EXPENSES OFFICE USE ONLY Kern (Describe the item purchased) Total cost (Exclude sales fox) OFFICE USE ONLY Item (Describe fhe item purchased) Total cost (E.clude sales tax) Dollars |Cents Dollars , Cents ~ 21 055 Cleaning, Laundry Supplies, and Paper Products -21 089 Housewares, Furnishings, Hardware, and Garden Supplies (mirrors, light bulbs, nails, etc 1 A S A $ B B C c D D E E F F G G Newspapers, Boohs, Postage, Stationery, and School Supplies Tobacco and Smoking Supplies H s H S J J K K ~ 21 063 Psrsonal Care, Drugs, and Medical Supplies (Ind'cate if prescribed by a doctor) - 21 097 Gas, Oil. Tolls, Parking Fees, and Other Vehicle Expenses; Public Transportation Expenses A $ A $ B j B C C D ~- 31 104 Clothing, Shoes, Jewelry, and Linens E casual slacks, etc ) F D $ C E H F J G Laundry and Diaper Service, Beauty and Barber Shop, Household Help, Babysitters Rent, Utilities, Fuel, Phone, Insurance (Specify period covered by expense, such as **eek, month, year) I K $ H s L 1 M K N L ! P ~21 113 Othtr Purchases and Enpensts (toys, hobbies, contributions, etc.) Q R M s Movies, Plays, Other Entertainment, Club and Other Membership Dues N P s S a T R U s 1 GIFTS t OFFICE USE ONLY Item (Describe the item purchased) Was this purchased for a member of this household' (Mark one) Total cost (Exclude sales fas') - Jl D/V Yes ] No Dollars [Cents A 1 ]2 $ B ' I 2 C ' I 2 D 1 J2 CE- 105 (4-3-731 APPENDIX C Listing of Detailed Sections of Questionnaire for Quarterly Panel Section 1— Household Record and Consumer Unit Determina- tion (listing of household members, their characteristics, and assignment to consumer units, if more than one). Section 2— Rented Living Quarters (determination of rental costs and facilities included in rent). Section 3— Owned Living Quarters and Other Owned Real Estate (recording of real property owned by respondents, for what purposes used, original cost, and related items). Section 4— Mortgage Payments and Ownership Costs (type and amount of indebtedness on each property, mortgage pay- ments, real estate taxes, and other ownership costs during sur- vey year). Section 5— Expenditures for Repairs, Alterations, and Main- tenance of Owned Property (quarterly inquiry on expenditures for these purposes for owned property). Section 6— Utilities, Fuels, and Household Help (quarterly inquiry on expenditures for these purposes). Section 7— Clothing and Household Textiles (quarterly inquiry on expenditures for clothing or footwear, clothing accessories, and jewelry; for alterations or repairs to clothing on footwear; for rental of clothing; for household linens; and for sewing materials). Section 8— Major Equipment Items— Inventory and Purchases (initial inventory of major household appliances and equipment and annual inquiry on expenditures during survey year for pur- chase of these items). Section 9— Minor Equipment Items— Inventory and Purchases (initial inventory of minor household appliances and semi- annual inquiry on expenditures for purchase of these items). Section 10— Service Contracts (initial listing of service con- tracts for maintenance or repair of equipment or appliances or for such services as pest control and lawn maintenance, and inquiry on annual expenditures for such contracts). Section 11 — Equipment Repairs (quarterly inquiry on expen- ditures not covered by service contracts for repairs to equip- ment and appliances). Section 12— Home Furnishings and Related Household Items (semiannual inquiry on expenditures for furniture and other furnishings such as floor coverings, furniture or window coverings, dinnerware, cookware, flatware; suitcases, luggage; decorative items; and furniture rental or repair). Section 13— Automobiles and Other Vehicles— Inventory or Purchases (initial inventory of automobiles and other vehicles and annual inquiry on expenditures for purchase of vehicles and for installment payments). Section 14— Renting and Leasing of Vehicles (inquiry— semi- annual in 1972, annual in 1973— on expenditures for these pur- poses). Section 15— Vehicle Operating Costs (quarterly inquiry on ex- penditures for tires and tubes, vehicle accessories, services and repair work, gasoline and oil, other current maintenance, and other operating costs such as license fees and parking). Section 16— Out-of-Town Trips and Vacations (quarterly inquiry on overnight trips and vacations with detailed expen- ditures for transportation, housing, food, and other purposes on each trip). Section 17— Insurance Other Than Health (listing of policies held and annual inquiry on expenditures by family for these purposes). Section 18-Hospitalization and Health Insurance (listing of policies held, type and coverage of policies, and annual inquiry on expenditures by family for these purposes). Section 19— Medical and Health Expenditures (semiannual in- quiry on out-of-pocket expenditures for hospital, physician, dental, and other professional services, and for eye examination, prescription drugs, and medical supplies). Section 20— Education Expenses (semiannual inquiry on out-of- pocket expenditures for tuition, fees, housing and living costs, purchase or rental of books and equipment, and other education costs, by type of school). Section 21 -Subscriptions and Memberships (listing of subscrip- tions for newspapers, magazines, and other reading or cultural activities; memberships in social, fraternal, and other nonbusi- ness organizations; and annual inquiry on expenditures for these purposes). Section 22— Miscellaneous Expenses (semiannual inquiry on ex- penditures for various large miscellaneous items such as wed- dings and other affairs, funeral and related costs, pet costs other than food, and moving expenses). Section 23-Expenses Patterns for Food, Beverages, and Other Selected Items (quarterly inquiry to obtain overall expenditures 58 59 by family for food and beverages and for a few other items such as garden supplies, phonograph records, film, and admissions to entertainment places). Section 24— Expenses Patterns for Selected Services and Goods (annual inquiry to develop overall expenditures by family for other smaller categories such as laundry, dry cleaning, public transportation, tobacco, and hair care). Section 25— General Housing and Consumer Unit Information (general description of housing units occupied by survey fami- lies and selected characteristics of family such as education of family head, and work experience and income of family mem- bers in the year preceding the survey year).' Section 26— Work Experience and Income in Survey Year and Other Selected Items (detailed work experience and income of family members in survey year and information on taxes, con- tributions, occupational expenses, charges in assets and liabili- ties during year, and credit finance charges). Information on income and work experience in the year preceding the survey year was obtained at the second quarterly visit so that there would be some information on these subjects for families which moved out of sample addresses before the detailed inquiry on these matters obtained in Section 26 at the final (fifth) quarterly interview. W.S. GOVERNMENT PRINTING OFFICE: 1978 281-049/1145 1-3 working papers " The titles listed below are available in printed copies or on microfiche. The date in parentheses following the title indicates year of publication. Copies of recent papers may be reviewed and/or ordered at district offices of the Department of Commerce located in principal cities throughout the United States. Additional information, including prices, may be obtained by writing to Subscriber Services (ASD), Bureau of the Census, Washington, D.C. 20233. No. 1. No. 2. No. 3. No. 4. No. 5. No. 6. No. 7. Raw Materials in the United States Economy: 1900-52 (Preliminary) (1954) The Role of the 1954 Census of Manufactures in Overcoming Problems of Industry Data (1956) Tract Data Compared for a 25-Percent Sample and a Complete Census (1956) Sampling in the 1950 Census of Population and Housing (1956) Occupational Trends in the United States: 1900 to 1950(1958) Raw Materials in the United States Economy (1964) Papers Presented at the Census Tract Conference, December 29, 1958(1959) No. 8. Materials on the Preparation and Conduct of the U.S.S.R. All-Union Population Census of 1959 (1959) No. 9. Historical Comparability of Census Manufactures Industries: 1929-1958 (1959) No. 10. Papers Presented at the Census Tract Conference, December 29, 1959 (1960) No. 11. Papers Presented at the Census Tract Conference, August 25, 1960(1960) No. 12. Papers Presented at the Census Tract Conference, August 28, 1961 (1962) No. 13. Papers Presented at the Census Tract Conference, September8, 1962 (1962) No. 14. The Spectral Analysis of Economic Time Series (1963) No. 15. Methodology and Scores of Socioeconomic Status (1963) No. 16. Procedural Report on the 1960 Censuses of Population and Housing (1963) No. 17. Papers Presented at the Census Tract Conference, September 5, 1963 (1964) No. 18. The Measurement of Performance Potential in Manu- facturing Establishments (1965) No. 19. Tests of Use of Post Office Resources to Improve Coverage of Censuses (1965) No. 20. Industry Classification and Sector Measures of In- dustrial Production (1965) No. 21. A Spectral Study of "Overadjustment" for Sea- sonality (1965) No. 22. Papers Presented at the Census Tract Conference, December 29, 1964 (1965) No. 23. Spectral Analysis and Parametric Methods for Sea- sonal Adjustment of Economic Time Series (1965) No. 24. Self-Enumeration as a Method for the 1970 Census of Housing (1966) No. 25. Measuring the Quality of Housing (1967) No. 26. Changes in the Structure of Manufacturing Employ- ment (1968) No. 27. Methodology of Consumer Expenditures Survey (1968) No. 28. Metropolitan Area Definition (1968) No. 29. Survey Applications of Social Psychological Ques- tions (1969) No. 30. Raw Materials in the United States Economy: 1900-1966(1970) No. 31. Price Variation in New FHA Houses: 1959-1961 (1971) No. 32. Pretests and Dress Rehearsals of the 1970 Census of Population and Housing: A Procedural History (1972) No. 33. An Estimate of A Quasi-Stable Age-Sex Distribution for Ghana in 1960(1972) No. 34. Investigation of Census Bureau Interviewer Char- acteristics, Performance, and Attitudes: A Summary (1972) No. 35. Raw Materials in the United States Economy: 1900-1960(1972) No. 36. Response Variance in the Current Population Survey (1972) No. 37. Who's Home When (1973) No. 38. Economic Censuses of the United States: Historical Development (1973) No. 39. Population and Housing Inquiries in U.S. Decennial Censuses, 1790-1970 (1973) No. 40. Studies in Occupational and Industrial Classification (1974) No. 41. Information Privacy and Statistics: A Topical Bibli- ography (1978) Superintendent of Documents U.S. Government Printing Office Washington, D.C. 20402 Official Business Postage and Fees Paid PENN STATE UNIVERSITY LIBRARIES bE33fl ADDDD7D 1 L,um-£.vz. technical papers The titles listed below are available in printed copies or on microfiche. The date in parentheses following the title indicates year of publication. Copies of recent papers may be reviewed and/or ordered at district offices of the Department of Commerce located in principal cities throughout the United States. Additional infor- mation, including prices, may be obtained by writing to Subscriber Services (ASD), Bureau of the Census, Washington, D.C. 20233. No. 1. The Sample Survey of Retail Stores: A Report on Methodology (1953) Description of the Sample for the Monthly Retail Trade Report (unnumbered revision) (1955) No. 2. Accuracy of Census Statistics With and Without Sampling (1960) Farm Population, 1880-1950 (1960) The Post Enumeration Survey: 1950 (1960) Tests and Revisions of Bureau of the Census Methods of Seasonal Adjustments (1961) The Current Population Survey Reinterview Program, Some Notes and Discussion (1963) The Current Population Survey, A Report on Methodology (1963) Trends in the Income of Families and Persons in the United States: 1947-1960 (1963) Reconciliation of the 1958 Census of Retail Trade with the Monthly Retail Trade Report (1963) Population Trends in the United States: 1900 to 1960 (1964) Response Errors in Collection of Expenditures Data by Household Interviews: An Experimental Study (1965) Estimating Trading-Day Variation in Monthly Economic Time Series (1965) Sampling Application in Censuses of Population and Housing (1965) The International Standard Industrial Classification and the U.S. Standard Industrial Classification (1965) The X-1 1 Variant of the Census Method II Seasonal Adjustment Program (1967) Present Value of Estimated Lifetime Earnings (1967) Trends in the Income of Families and Persons in the United States: 1947 to 1964 (1967) Changes Between the 1950 and 1960 Occupation and Industry Classification (1968) The Current Population Survey Reinterview Program, January 1961 through December 1966 (1968) Correlation Between United States and International Standard Industrial Classifications (1969) Characteristics of America's Engineers and Scientists: 1960 and 1962 (1969) No. 3. No. 4. No. 5. No. 6. No. 7. No. 8. No. 9. No. 10. No. 11. No. 12. No. 13. No. 14. No. 15. No. 16. No. 17. No. 18. No. 19. No. 20. No 21. No. 22. Measures of Overlap of Income Distributions of White and Negro Families in the United States (1970) No. 23. The Position of United States in World Commodity Exports in 1968 (1970) No. 24. The Annual Survey of Manufactures: A Report on Methodology (1971) No. 25. Demographic Computer Library (1971) No. 26. 1970 Occupation and Industry Classification System Showing Sources (1972) No. 27. Sampling Applications of the 1970 Census Publication, Maps, and Public Use Summary Files (1972) No. 28. From the Old to the New 1972 SIC for Establish- ments (1972) No. 29. Scientific and Technological Development Activities of the Census (1973) No. 30. Census County Division, Past and Future (1973) No. 31. Consistency of Reporting Ethnic Origin in the Cur- rent Population Survey (1974) No. 32. Standards for Discussion and Presentation of Errors in Data (1974) No. 33. Characteristics of Persons in Engineering and Scientific Occupations: 1972 (1974) No. 34. indexes to Survey Methodology Literature (1974) No. 35. Family (Money) Income: 1947 to 1971 : Summarizing Twenty-Five Years of a Summary Statistic (1974) No. 36. Census Bureau Programs to Measure Consumer Pur- chase Expectations: 1959-1973 (1974) No. 37. The Census Bureau: A Numerator and Denominator for Measuring Change (1975) No. 38. Comparison of Persons of Spanish Surname and Per- sons of Spanish Origin in the United States (1975) No. 39. Guide for Local Area Population Projections (1977) No. 40. The Current Population Survey: Design and Method- ology (1978) No. 41. An Evaluation of 1970 Census Occupational Classi- fication (1978) No. 42. 1976 Survey of Institutionalized Persons: Methods and Procedures (1978) No. 43. Graphic Presentation of Statistical Information (1978) No. 44. Standard Statistical Establishment Program (1978) No. 45. The 1972-73 U.S. Consumer Expenditure Survey: A Preliminary Evaluation (1978)