PDF (1036 K) - National Bureau of Economic Research

This PDF is a selection from an out-of-print volume from the National Bureau of Economic Research

Volume Title: Source Book of Statistics Relating to Construction

Volume Author/Editor: Robert E. Lipsey and Doris Preston

Volume Publisher: NBER

Volume ISBN: 0-87014-082-5

Volume URL: http://www.nber.org/books/lips66-1

Publication Date: 1966

Chapter Title: Source Notes and Descriptions

Chapter Author: Robert E. Lipsey, Doris Preston

Chapter URL: http://www.nber.org/chapters/c1592

Chapter pages in book: (p. 227 - 289)

III

SOURCE NOTES AND DESCRIPTIONS

A. Construction Contracts

DODGE CONSTRUCTION CONTRACT SERIES, MONTHLY

General Description

SOURCE: These series are based on informationcompiled by the F. W. Dodge Corporation ofNew York (now the F. W. Dodge Company).Monthly data for years prior to 1942 weretaken directly from the files of the DodgeCorporation. Beginning with 1942 the basicsources have been two monthly bulletins: (1)Construction Contracts Awarded, 37 EasternStates (1942—1956); and (2) Construction Con-tracts, United States Summary (1957—1963).The NBER classification by type of construc-tion differs, in some cases, from the Dodgeclassification. These rearrangements are notedin the descriptions of the individual series.With the exception of the three series whichhave been discontinued, current data for theseries presented here are available in Cons truc-tion Contracts, U.S. Summary. In some casesthe titles of the NBER and Dodge series differ,and in other cases the NBER series are combi-nations of two or more Dodge series. (See in-dividual series descriptions for specific Dodgesources.) Current data on the value series fortotal, public, and private construction; forresidential and nonresidential building; andfor public works and utilities are also availablemonthly in the Survey of Current Business(U.S. Department of Commerce).

Annual data are from the same sourcesand, for 1932—44, from Record of ContractsAwarded, 37 Eastern States. In addition tothe series on floor space given here, the Dodgesources also include floor space data on one-and two-family dwellings, apartment houses,and hotels.

The Dodge "construction contracts" statis-tics are conceptually a measure of "commit-

ments of work about to start." 1 They includeconstruction projects for which there are nogeneral contracts, as well as those for whichsuch contracts are awarded. In general, proj-ects are included in the statistics when it isreasonable to expect construction to startwithin sixty days.

The data include new construction, addi.tions, and major alterations, and excluderepairs and maintenance. Force-account workis included only when executed with materialsearmarked for specific projects at time of pur.chase. All urban and rural nonfarm areas areincluded, while all but a negligible part offarm building is excluded. Both private andpublic construction are covered, except forprivately owned projects below a minimumvalue. The minimum valuation, currently$10,000, has varied considerably, falling aslow as $500 during the early 1930's. Publiclyowned projects are included without mini-mum, except for minor deviations. In themost active counties of the eleven westernstates (forty-eight-state segment), two addi-tional exclusions apply: privately owned addi-tion and alteration projects valued at lessthan $100,000, and separately built residentialgarages.

The types of projects included in eachNBER series are enumerated in the descrip-tion of the individual series. Structures areclassified according to function. Combinationprojects are classified according to the primepurpose for which the building is constructed,insofar as can be determined. Additions andalterations to existing buildings are classifiedin accordance with the use to which theremodeled structure is to be put.

1 Composition of Dodge Construction Contract Statis-tics, F. W. Dodge Corporation, January 1968.

230 SOURCE NOTES AND DESCRIPTIONS

Valuation figures represent, as nearly aspossible, actual construction costs, includingsubcontracts for such items as plumbing, heat-ing, electrical work, roofing, and normalconnecting utilities, and excluding land andarchitects' fees. Cost of industrial equipmentwhich is not an integral part of the structureis excluded, except for special purpose equip-ment in petroleum refineries, outdoor chemi-cal plants, electrical generating, power, andheating plants, and water and sewage treat-ment plants.

Floor space figures represent footage underroof, exclusive of basement. Where buildingpermit data are the basis of the statistics,floor area is estimated from construction costs,applying local building cost differentials ofthe Dow Building Cost Calculator tonationally established cost-per-square-footrates.

Geographic coverage has been increased inseveral steps since the Dodge reporting pro-gram began. The earliest Dodge statisticscover total construction in the New Englandstates, beginning in 1901. Data coveringtwenty-seven northeastern states and the Dis-trict of Columbia are available from 1910 fortotal construction, from for residentialbuildings, and from 1919 for all other cate-gories of construction. The addition of ninesouthern states to the reporting programbetween 1920 and 1923 brought the total tothirty-six. Texas was added in 1924, makinga total of thirty-seven states. In January 1956the statistical coverage was expanded toinclude eleven western states, and the statisticshave covered forty-eight states since that time.Alaska and Hawaii have not been included inthe reporting program. Because of thesechanges in geographic coverage, the series arepresented here in separate segments for twenty-seven, thirty-six, thirty-seven, and forty-eightstates, with overlaps of one or more yearsbetween segments. (Early data on total con-struction in the New England states are shownas a separate -series.)

Until 1956 the statistics developed by the

Dodge Corporation on value, floor space, andnumber of projects for various categories ofconstruction were a by-product of a daily con-struction news service known as "DodgeReports." Statistical summaries were basedon information collected by news reporters,through reading and clipping services, andthrough direct mail inquiries. The reportersgathered information primarily from archi-tects, designers, engineers, contractors, build-ers, and owners; and supplemented this byconsulting building inspectors and other gov-ernment sources. This procedure is still gen-erally followed.

Two new procedures were introduced in1956—57: (1) Estimates for the eleven westernstates added in 1956 are based primarily onbuilding permits, supplemented by newssources and by sampling. These states are notcovered by the construction news service, andbuilding permits are more universally requiredthan in the thirty-seven eastern states. (2) Thebasic source of information on privatelyowned one- and two-family houses was changedfrom "Dodge Reports" to building permits.For one group of counties, estimated toaccount for 85 per cent of one- and two-familyhousing starts, estimates are based on completereporting of building permits from the mostactive permit-issuing places, reports on per-mits from 75 per cent of less active permit-issuing places, and, for the eleven westernstates, estimates by reporters of buildingactivity in nonpermit areas. In the secondgroup of counties, estimated to account for 15per cent of one- and two-family construction,activity is projected from permit data for asample of counties in the first group, usinginformation on the climatic and economiccharacteristics of each county. An estimated 5per cent of total private one- and two-familyhouses in all counties is excluded. The housesnot covered are those in nonpermit-issuingareas in the eastern -states and those in lessactive nonreporting permit-issuing areas.2

2 Ibid.

CONSTRUCTION CONTRACTS 231

Both new procedures applied to the entireforty-eight-state segment. Revisions based onthe new method of estimating one- and two-family houses were also developed for thethirty-seven-state segment back to 1947, forthe monthly series on one- and two-familyhouses, total residential construction, andtotal construction. These revisions made thethirty-seven-state segment for 1947—56 com-parable in coverage (except for geographicarea) to the forty-eight-state segment, but notcomparable to the earlier data for thirty-sevenstates. The data presented in both the annualand monthly sections of this volume for seriesA5, Total Construction, Value; A7, TotalResidential, Value; A9, One- and Two-FamilyHouses, Value; and A8, Total Residential,Floor Space are the unrevised figures for 1947—56, which are comparable to the earlier thirty-seven-state data. The revised annual figuresare shown in a footnote to the Annual DataTable (Section lI-A). Series A31, Number ofNew Dwelling Units Provided, is presentedonly on the revised basis, and only from 1947on. The revisions in the data on one- andtwo-family houses were not incorporated byDodge in their series on total private construc-tion.

Further improvements in the samplingmethods for one-family houses were put intoeffect in January 1958.

The handling of corrections and revisionsin data was modified in 1958. Before that date,adjustments were made in the months duringwhich they were ascertained rather than in themonths to which they applied. Since March1958, the procedure has been to handle onlyupward adjustments in this way, and to enterdownward adjustments only in the cumulativeto-date statistics.3

The degree to which the Dodge data coverall construction commitments in the UnitedStates has been estimated'for certain categories

a Adjustments which were made only in the cumula-tive figures account for the discrepancies, beginning1958, between the monthly and annual series shownhere.

and areas. In 1938 the Department of Com-merce estimated that the Dodge statisticscovered "as much as 80 to 90 per cent of allnew nonresidential building constructionregardless of the size of the project" in thethirty-seven eastern states then included inthe Dodge program.4 A report to the Bureauof the Budget in 1958 suggested that underre-porting in the thirty-seven eastern states wasat that time about 10 per cent of total valua-

In 1963, the Dodge Corporation estimatedthat 95 per cent of all privately owned, per-manent, nonfarm, one- and two-family housesstarted in the forty-eight states were accountedfor in the Dodge statistics.° Types of con-struction for which coverage tends to besmallest are construction in rural areas, force-account projects, additions and alterations,and smaller projects, particularly one- andtwo-family houses prior to 1957.

The Dodge construction contract data forthe thirty-seven eastern states are the basis forthe Bureau of the Census estimates of valueof new construction put in place for all cate-gories of private nonfarm nonresidential con-struction except public utilities.

For further description and evaluation ofthe Dodge series, see Composition of DodgeConstruction Contract Statistics, January 1963;Geoffrey H. Moore, Business Cycle Indicators,Volume II, Princeton University Press forNBER, 1961, pp. 12—15; Elmer Clark Bratt,Appraisal of Statistical In formation on Con-struction in the U.S., A Report to the BudgetBureau (together with Appendices referred toin footnote 5); Chawner, Construction Activityin the U.S. (see footnote 4).

Original monthly data were seasonally

4 Lowell J. Chawner, Jr., Construction Activity in theUnited States, 1915—37, Domestic Commerce Series No.99, Bureau of Foreign and Domestic Commerce, U.S.Department of Commerce, 1938, pp. 45—46.

5 Elmer Clark Bratt, Appendices to Appraisal ofStatistical Information on Construction in the U.S.,September 5, 1958, p. 24.

6 Composition of Dodge Construction ContractStatistics, January 1963.


adjusted by the National Bureau. For mostseries the adjusted data are shown in segmentscorresponding to those for which the originaldata were available. However, for ten series,the adjusted data are shown as a continuousseries, all raised to the level of the forty-eight-state data. For these ten series, raising ratioswere applied to the original data before theseasonal adjustment was computed. Ratiosused for linking segments were based on thetwo annual totals for the earliest year of theoverlap period.

Individual Series

Al. Total Construction, New England States,Value, 1901—1918.SOURCE: Dodge Corporation files. This

series was discontinued by NBER when datafor a larger area became available. However,regional data are compiled and published bythe Dodge Corporation.

CONTENT: See general description of Dodgeseries.

A2. Total Construction, Value, 1910—1963.SOURCE, Monthly data:

1910—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1. 1956 (48states): Record of Construction Contracts, 48States (1956 summary). 1957—63: ConstructionContracts, United States Summary. Currentdata are also available in Survey of CurrentBusiness.

Annual data: Same as above, except 1932—1944: Record of Contracts Awarded, 37 East-ern States.

CONTENT: See general description of Dodgeseries.

A3. Total Private Construction, Value, 1919—1963.NBER Definition*SOURCE:

1919—41: Dodge Corporation files. 1942—56:Construction Contracts Awarded, 37 Eastern

States, Section I, Table 1, "Total ResidentialBuildings," "Commercial Buildings," "Man-ufacturing Buildings," and "Utilities (Publicand Private)"; and Section II, Table 6,"Theatre Bldgs., Etc.," "RR-Bus-Air Termi-nal Bldgs.," minus "Water Supply Systems."Current data for some components not avail-able. Series discontinued.

CONTENT: Residential buildings (SeriesA7), Commercial and Industrial building(Series Al4), Public Utilities (Series A28).

Dodge Definition *SOURCE:

1932—44: Record of Contracts Awarded, 37Eastern States. 1945—56 (37 states): Construc-tion Contracts Awarded, 37 Eastern States, Sec-tion I, Table 5. 1956—63 (48 states): Construc-tion Contracts, United States Summary, tableon "Public vs. Private. Ownership." Currentdata are also available in Survey of CurrentBusiness.

CONTENT: Building and nonbuilding proj-ects constructed under private ownership.

A4. Public Construction, Value, 1919—1963.NBER Definition *SOURCE:

1919—41: Dodge Corporation files. 1942—56(37 states): Construction Con tracts A warded,37 Eastern States, Section I, Table 1, "Educa-tional & Science Bldgs.," "Hospital & Institu-

* Series AS and A4 were each compiled according totwo different concepts, designated as "NBER definition"and "Dodge definition." The NBER definitions ofprivate and public were based on type of construc-tion project. These series, which include specificcategories of projects, were compiled for the period1919—56, after which data for some components werenot available. The two series do not add to total con-struction contracts, since a few minor categories werenot allocated to either private or public. The Dodgedefinitions of private and public are based on owner-ship. Beginning 1932 a breakdown by ownership wasavailable and series on privately owned and publidyowned construction contracts were therefore compiledfor 1932—63. The data are presented. here separatelyfor each of the two definitions. For purposes of sea-sonal adjustment and cyclical analysis, they were com-bined, using the Dodge definition for 1932—63 andextrapolating back to 1919 by the NBER definition,linking by the ratio of the two 1932 annual totals.


tional Bldgs.," "Public Buildings," and "Pub-lic Works"; and Section II, Table 6, "Gym-nasium & Athletic Bldgs.," and "Water SupplySystems," minus "Parks-Playgrounds, Etc."Current data for some components not avail-able. Series discontinued.

CONTENT: (1) Public buildings (SeriesA22); and (2) public works (Series A26).

Dodge Definition *SOURCE:

1932—44: Record of Contracts Awarded, 37Eastern States. 1945—56 (37 states): Construc-tion Contracts Awarded, 37 Eastern. States,Section I, Table 5. 1956—63 (48 states): Con-struction Contracts, United States Summary,table on "Public vs. Private Ownership."Current data are also available in Survey ofCurrent Business.

CONTENT: Building and nonbuilding proj-ects constructed under public ownership.

A5. Total Building Excluding Public Worksand Utilities, Value, 1919—1963.

A6. Total Building Excluding Public Worksand Utilities, Floor Space, 1919—1963.SOURCE:

1919—41: Dodge Corporation files. 1942—56(37 States): Construction Contracts Awarded,37 Eastern States, Section I, Table 1, "TotalBuilding"; and Section II, Table 6, "Parks-Playgrounds, Etc.," and "Pipe Lines—Oil &Gas Wells," minus "RR-Bus-Air TerminalBldgs." 1956—63 (48 states): Construction Con-tracts, United States Summary, "Total Build-ing." Current data for value series are alsoavailable in Survey of Current Business (sumof "Residential" and "Nonresidential" build-ing).

CONTENT: The 27-, 36-, and 87-state seg-ments include all construction except publicworks as defined in Series A26, and publicutilities as defined in Series A28. The 48-statesegment includes the same, plus railroad, bus,and air terminal buildings, minus parks andplaygrounds, pipe lines, and oil and gaswells.

A7. Total Residential Buildings, Value,1915—1963

A8. Total Residential Buildings, Floor Space,1919_1963.aSOURCE, Monthly data:

1915—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1. 1956 (48states): Record of Construction Contracts, 48States (1956 summary). 1957—63: ConstructionContracts, United States Summary. Currentdata for value series are also available in Sur-vey of Current Business.

Annual Data: Same as above except 1932—44: Record of Contracts Awarded, 37 EasternStates.

CONTENT: One- and Two-Family Dwellings(Series A9), Apartment Houses (Series AlO),and Hotels (Series All), and other shelter,including boarding and rooming houses,tourist cottages and camp buildings.

A9. and Two-Family Dwellings, Value,1919_1963.aSOURCE, Monthly data:

1919—41: Dodge Corporation files. 1942—56(37 states): Construction Contracts Awarded,37 Eastern States, Section I, Table 1. 1956(48 states): Record of Construction Contracts,48 States (1956 summary). 1957—63: Construc-tion Contracts, United States Summary.

Annual Data: Same as above, except 1932—44: Record of Contracts Awarded, 37 EasternStates.

CONTENT: All houses containing one ortwo dwelling units built for owner occupancy,sale or rent.

AlO. Apartment Houses, Value, 1919—1963.SOURCE:

1919—42: Dodge Corporation files. 1943—56(37 states): Construction Contracts Awarded,

• See note on p. 232.a See general description of Dodge series for 1957

revision in method of compiling statistics on one- andtwo-family houses.


37 Eastern States, Section I, Table 1, "Apart-ment Buildings" and "Dormitories." 1956—63(48 states): Construction Contracts, UnitedStates Summary, "Apartment Buildings" and"Dormitories."

CONTENT: (1) Apartment buildings (allbuildings designed to house families and con-taining three or more dwelling units); (2)dormitories of schools or institutions, nursesor staff homes, fraternity and sorority houses,etc.; (3) apartment hotels, designed primarilyto house families on a relatively permanentbasis (included 1919-42 only; thereafterincluded in "Hotels," Series All).

All. Hotels, Value, 1919—1962.SOURCE:

1919—42: Dodge Corporation files. 1943—56(37 states): Construction Contracts Awarded,37 Eastern States, Section I, Table 1, "Hotelsand Apartment Hotels" through September1949, thereafter "Hotels." 1956—62 (48 states):Construction Contracts, United States Sum-mary. Current data: Data on hotels not avail-able separately. Effective January 1963,motels and hotels are combined into oneclassification, since the distinction between thetwo has lost meaning. Combined data in Con-struction Contracts, United States Summary.

CONTENT: (1) All transient hotels, inns,etc.; (2) apartment hotels designed primarilyto house families on a relatively permanentbasis (included only from 1943 on; prior tothat date included in "Apartment Houses,"Series A 10).

A 12. Total Nonresidential Buildings, Value,1919—1963.

A13. Total Nonresidential Buildings, FloorSpace, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56(37 states): Construction Contracts Awarded,37 Eastern States, Section I, Table 1, "Total

Non-Residential Bldg."; and Section II, Table6, "Parks-Playgrounds, Etc.," and "Pipe Lines—Oil & Gas Wells," minus "RR-Bus-AirTerminal Bldgs." 1956 (48 states): Record ofConstruction Contracts, 48 States (1956 sum-mary). 1957—63: Construction Contracts,United States Summary. Current data forvalue series are also available in Suruey ofCurrent Business.

CONTENT: (1) Commercial buildings (SeriesA16 and A17); (2) industrial buildings (SeriesA18 and A19); (3) public and institutionalbuildings (as defined in Series A20 and A21,except that theater buildings are includedthroughout); and (4) miscellaneous nonresi-dential buildings (railroad, bus, and air termi-nal buildings included only for 48-state seg-ment).

A14. Total Commercial and Industrial Build-ings, Value, 1919—1963.

A15. Total Commercial and Industrial Build-ings, Floor Space, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56(37 states): Construction Contracts Awarded,37 Eastern States, Section I, Table 1, "Com-mercial Buildings" and "Manufacturing Build-ings"; and Section II, Table 6, "TheatreBldgs., Etc." and "Pipe Lines—Oil & GasWells." 1956 (48 states): Record of Construc-tion Contracts, 48 States (1956 summary).1957—63: Construction Contracts, UnitedStates Summary, "Commercial Bldgs." plus"Manufacturing Bldgs."

CONTENT: (1) Commercial buildings (SeriesAl6 and A17); (2) industrial buildings (SeriesA18 and A19); and (3) theater buildings:theaters, music conservatories, radio broad-casting studios, etc. (Theater buildingsincluded only in 27-, 36-, and 37-state series;excluded in 1956—63 segment for 48 states.)The floor space series does not cover pipe linesand oil and gas wells.


A 16. Commercial Buildings, Value, 1919—1963.

A17. Commercial Buildings, Floor Space,1919—1963.

SOURCE, Monthly data:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1. 1956 (48states): Record of Construction Contracts, 48States (1956 summary). 1957—63: ConstructionContracts, United States Summary.

Annual data: Same as above, except 1932—44: Record of Contracts Awarded, 37 EasternStates.

CONTENT: Public garages, auto servicestations (gasoline, battery, greasing, etc.), air-craft hangars, etc.; banks, building and loanassociations, and other financial buildings;office and loft buildings; all retail stores andshops; all restaurant buildings; all commer-cial warehouses and storage buildings, exclud-ing cold storage, grain elevator and storagesilos. Airports were excluded beginning in1931. (Their total value amounted to less than2 per cent of total commercial buildings in1930.)

A18. Industrial Buildings, Value, 1919—1963.

A19. Industrial Buildings, Floor Space, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1, "Manufac-turing Buildings," and Section II, Table 6,"Pipe Lines—Oil & Gas Wells." 1956 (48states): Record of Construction Contracts, 48States (1956 summary). 1957—63: ConstructionContracts, United States Summary, "Manufac-turing Bldgs."

CONTENT: Manufacturing, assembly andwarehouse buildings in processing or mechani-cal industries; refrigeration, ice and cold stor-age plants of warehouses; industrial grainelevators and storage silos; dry cleaning plants,

laundries, and light manufacturing buildings.Pipe lines and oil and gas wells are includedexcept for 1956—63, 48-state segment. Indus-trial power plants are excluded throughout.The floor space series does not cover pipe linesand oil and gas wells.

A20. Total Public and Institutional Buildings,Value, 1919—1963.

A21. Total Public and Institutional Buildings,Floor Space, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1, "Educa-tional & Science Bldgs.," "Hospital and Insti-tutional Bldgs.," "Public Buildings," "Reli-gious Buildings," and "Social & RecreationalBldgs."; and Section II, Table 6, "Parks-Play-grounds, Etc.," minus "Theatre Bldgs., Etc."1956 (48 states): Record of Construction Con-tracts, 48 States (1956 summary), "Educa-tional," "Hosp.-Institutional," "Public Bldgs.,""Religious," and "Social-Recreational." 1957—63: Construction Contracts, United StatesSummary, "Educational & Science Bldgs.,""Hospital Bldgs." (since 1960 called "HospitalSc Institutional Bldgs."), "Public Bldgs.," "Re-ligious Bldgs.," and "Social & RecreationalBldgs."

CONTENT: Public Buildings (Series A22)and Institutional Buildings (Series A24). Thefloor space series does not include parks andplaygrounds.

A22. Public Buildings, Value, 1919—1963.

A23. Public Buildings, Floor Space, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1, "Educa-tional & Science Bldgs.," "Hospital & Institu-tional Bldgs.," and "Public Buildings"; and


Section II, Table 6, "Gymnasium & AthleticBldgs." 1956 (48 states): Record of Construc-tion Contracts, 48 States (1956 summary),"Educational," "Hosp.-Institutional," and"Public Bldgs." 1957—63: Construction Con-tracts, United States Summary, "Educational& Science Bldgs.," "Hospital Bldgs.," and"Public Bldgs."

coNTENT: Educational and commerciallaboratories, science buildings, observatories,planetariums; private and public libraries,museums, and fine arts buildings; all publicand private school and college buildings;hospitals, clinics, and other institutional build-ings; public administration buildings, postoffices, armories, arsenals, jails, penitentiaries;comfort stations, fire stations, zoo buildingsand other miscellaneous public buildings;military and naval buildings. Gymnasium andathletic buildings are included only from1938 to 1956 and only for the 37-state segment.

A24. Institutional Buildings, Value, 1919—1963.

SOURCE:

1919—41: Dodge Corporation files. 1942—56 (37states): Construction Con tracts Awarded, 37Eastern States, Section I, Table 1, "ReligiousBuildings" and "Social and RecreationalBldgs."; and Section II, Table 6, "Parks-Play-grounds, Etc.," minus "Theatre Bldgs., Etc."and "Gymnasium & Athletic Bldgs." 1956 (48states): Record of Construction Contracts,48 States (1956 summary), "Religious" and"Social-Recreational." 1957—63: ConstructionContracts, United States Sunzmary, "ReligiousBldgs." and "Social 8c Recreational Bldgs."

CONTENT: Churches and other religiousbuildings; assembly buildings, auditoriums,and community houses; clubs and lodges; gym-nasium and athletic buildings (excluded 1938—56 for 37 states); parks, playgrounds, outdoorstadiums and similar projects (1938—56 for 37states only); miscellaneous social and recrea-tional buildings. Theatre buildings areincluded in 1956—63 data for 48 states, butexcluded in other segments.

A25. Public Works and Utilities, Value, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37Eastern States, Section I, Table 1, "PublicWorks (Public & Private)," "Utilities (Public& Private)"; and Section II, Table 6, "RR-Bus-Air Terminal Bldgs.," minus "Parks-Play-grounds, Etc." and "Pipe Lines—Oil & GasWells." 1956—63 (48 states): Construction Con-tracts, United States Summary, "Total HeavyEngineering Construction" (beginning Feb-ruary 1963 designated "Total Non-BuildingConstruction" with no change in content).Current data are also available in Survey ofCurrent Business (sum of "Public Works"and "Utilities").

CONTENT: 27-, 36-, and 37-state segments:(1) Public works (Series A26); and (2) publicutilities (Series A28). 48-state segment: Sameas above, plus parks and playgrounds, pipelines, and oil and gas wells, and minus rail-road, bus, and air passenger terminals.A26. Public Works, Value, 1919—1956.

SOURCE:1919—41: Dodge Corporation files. 1942—56:Construction Contracts Awarded, 37 EasternStates, Section I, Table 1, "Public Works(Public & Private)"; and Section II, Table 6,"Water Supply Systems," minus "Parks-Play-grounds, Etc." Current data for some com-ponents not available. Series discontinued.

CONTENT: Streets and highways, bridgesexcept those built by railroads or other pri-vate corporations, dams and reservoirs, water-front developments, sewerage systems andtreatment plants, water supply systems, canals,drainage ditches, incinerators, parking lots,memorials, and all other miscellaneous publicworks except parks and playgrounds.

A27. Streets, Roads, and Bridges, Value, 1919—1963.

SOURCE:1919—41: Dodge Corporation files. 1942—56 (37states): Construction Contracts Awarded, 37


Eastern States, Section II, Table 6, "Streets &Highways" and "Bridges." 1956—63 (48 states):Construction Contracts, United States Sum-mary, "Streets and Highways" and "Bridges."

CONTENT: (1) Streets, highways, alleys,vehicular tunnels; and (2) bridges except thosebuilt by railroads or other private corpora-tions.

A28. Public Utilities, Value, 1919—1956.SOURCE:

1919—41: Dodge Corporation files. 1942—56:Construction Contracts Awarded, 37 EasternStates, Section I, Table 1, "Utilities (Public& Private)"; and Section II, Table 6, "RR-Bus-Air Terminal Bldgs.," minus "Pipe Lines—Oil & Gas Wells," and "Water Supply Sys-tems." Current data for some components notavailable. Series discontinued.

CONTENT: Electric light and power plants,including industrial power plants; substationsand lines, radio stations, and conduit systems,etc.; gas plants and mains; railroad construc-tion, including bridges, tunnels, subways,crossing eliminations, etc. (excluding build-ings); airports (starting 1931); railroad, busand air passenger terminal buildings (starting1938); all other utility construction exceptpipe lines, oil and gas wells, and water supplysystems.

A29. Public Works and Utilities Under Pri-vate Ownership, Value, 1932—1963.

A30. Public Works and Utilities Under Pub-lic Ownership, Value, 1932—1963.

SOURCE:1932—41: Record of Contracts Awarded, 37Eastern States, tables on "Utilities—by Owner-ship" and "Public Works—by Ownership."1942—56 (37 states): Construction ContractsAwarded, 37 Eastern States, Section I, Table5. 1956—63 (48 states): Construction Contracts,United States Summary, table on "Public vs.Private Ownership."

cONTENT: Privately owned (Series A29)and publicly owned (Series A30) projects inthe following categories: streets and high-ways, bridges, dams and reservoirs, waterfrontdevelopments, sewerage systems, parks andplaygrounds, electric light and power plants,gas plants and mains, pipe lines, oil and gaswells, water supply systems, railroad construc-tion, airports (excluding buildings), mis-cellaneous public works and utilities.

A31. Residential Buildings, Number ofDwelling Units Provided, 1947_1963.a

SOURCE:1947—55: Record of Construction Contracts, 37States, Revised (annual summaries), "DwellingUnits—Number—New Building Only"; plusunits provided in apartment buildings "byconversion" from Construction Con tractsAwarded, 37 Eastern States, Section I, Table3. 1956 (37 states): Record of ConstructionContracts, 37 Eastern States, Revised (1956summary), "Dwelling Units—Number." 1956—63 (48 states): Construction Contracts, UnitedStates Summary.

CONTENT: Dwelling units provided in newor converted apartment buildings (all build-ings designed to house families and contain-ing three or more dwelling units), and in newone- or two-family houses. Converted projectsare additions or alterations to existing build-ings that create additional dwelling units.Dwelling units thus created in apartmentbuildings are included if the conversioninvolves expenditures above the $10,000 mini-mum valuation. Excluded are dwelling unitsprovided in converted one- or two-familyhouses, and in "combination projects"—build-ings that are primarily stores, office buildings,or other nonresidential structures. Alsoexcluded are trailers, mobile homes, temporarystructures, and summer bungalows withoutcooking facilities.

a Data for this series indude 1957 revisions on one- andtwo-family houses, for the period 1947—1956. (Seegeneral description of Dodge series.)


ENGINEERING CONSTRUCTION CONTRA C TS,ENGINEERING NEWS-RECORD, MONTHLY

General Description

SOURCE: Data are reprinted by permission ofEngineering News-Record, copyrighted, Mc-Graw-Hill Publishing Company, Inc., NewYork.

Data on value of contracts awarded forheavy engineering construction projects werecompiled by Engineering News-Record from1913 to 1962 from information gathered by itsfield staff and reported in Construction Daily.Monthly statistics, presented here in terms ofweekly averages, were derived from monthlysummary tables in Engineering News-Recordfor years 1943—1962, and from monthly totalsobtained directly from ENR files for yearsprior to 1943. The monthly series on total con-struction contracts was also carried in the Sur-vey of Current Business.

The statistics cover contracts and projectsannounced by owner-builders for public andprivate construction projects in the U.S. thatare valued over a specific minimum. Alaskaand Hawaii are included beginning in 1959.The minimum valuation varies for differenttypes of construction, and has been changedseveral times over the period covered, toreflect changes in construction costs. From1959 to 1962 it ranged from $53,000 for water-works, excavation, drainage, and irrigationprojects to $400,000 for nonindustrial build-ings (for details of minimum figures, seeindividual series descriptions). Because of theminimum valuation, a large segment of resi-dential building is excluded. In the residen-tial category, only larger housing projects,hotels, and apartment buildings are included.Data are available in the published source fortotal construction, for private and public con-struction by type of project, and for federalconstruction separately. A regional breakdownfor each type of construction is also given.Series are presented here on total construction

and four selected categories (for types of con-struction included in each, see individualseries descriptions).

The monthly figures from ENR cover four-or five-week periods. Weekly averages werecomputed by NBER by dividing monthlytotals by the number of weeks reported ineach month. The data were seasonally ad-justed by NBER.

The series on contracts awarded were dis-continued after 1962. However, the Engineer-ing News-Record has developed statistics onproposed new construction and backlog ofproposed construction. These data representanticipated construction at an earlier stagein the planning process than the contractaward series. One of these series, "new advanceplanning" for total heavy construction is avail-able in the Survey of Current Business, be-ginning with the March 1963 issue.

Individual Series

A32. Total Construction, Value, 1913—1962.Sum of A33, Total Building, and A34, Con-struction Other Than Building. (See below.)

A33. Total Building, Value, 1913—1962.Includes private buildings: industrial, commer-cial, and "mass housing"; and public build-ings: residential and other, except federalbuildings before 1932. Federal buildings werenot separable from federal nonbuilding con-struction for 1913—1931, and were includedin Series A34 for that period.

Minimum values of industrial buildingsincluded are shown in Series A35 below. Mini-mum values for all other buildings are asfollows: 1913 through June 1932, $150,000;July 1932 through 1935, $105,000; 1936, $140,-000; January 1937 through April 1946, $150,-000; May 1946 through November 1947, $205,-000; December 1947 through November 1950,$250,000; December 1950 through 1954, $300,-


000; 1955 through 1958, $344,000; 1959through 1962, $400,000.

A34. Construction Other Than Building,Value, 1913—1962.

Includes (1) public works: waterworks, sewer-age, bridges, earthwork, waterways, streets androads, and "unclassified"; and (2) private con-struction: bridges and "unclassified." For1913—1931, federal buildings, which were notseparable from other federal construction, arealso included.

Projects of the following minimum costsare included. Waterworks, excavation, drain-age, and irrigation: 1913 through June 1932,$15,000; July 1932 through 1935, $10,500;1936, $14,000; January 1937 through April1946, $15,000; May 1946 through November1947, $22,500; December 1947 through Novem-ber 1950, $28,000; December 1950 through1954, $34,000; 1955 through 1958, $44,000;1959 through 1962, $53,000. Other publicworks: 1913 through June 1932, $25,000; July1932 through 1935, $17,750; 1936, $23,000;

1937 through April 1946, $25,000; May 1946through November 1947, $40,000; December1947 through November 1950, $50,000; Decem-ber 1950 through 1954, $60,000; 1955 through1958, $73,000; 1959 through 1962, $88,000.

A35. Industrial Buildings, Value, 1913—1962.Includes private industrial buildings abovethe following minimum costs: 1913 throughJune 1932, $40,000; July 1932 through 1935,$28,500; 1936, $37,000; 1937 through April1946, $40,000; May 1946 through November1947, $55,000; December 1947 through Novem-ber 1950, $68,000; December 1950 through1954, $82,000; 1955 through 1958, $93,000;1959 through 1962, $110,000.

A36. Federal Construction, Value, 1913—1962.Includes contracts of the federal governmentfor construction of building or nonbuildingprojects which are above the minimum costfor the particular type of constructioninvolved. (See Series A33 and A34 for detailsof minimum costs.)

PORTLAND CEMENT ASSOCIATION, ANNUAL

A37. Concrete Pavement Contract Awards forRoads, Streets, and Alleys, Yardage,1919—1963.

SOURCE:1919—1934: Survey of Current Business, 1936Supplement. 1935—1938: Business Statistics,1959 Supplement to Survey of Current Busi-ness. 1939—1962: Business Statistics, 1963 Sup-plement to Survey of Current Business. 1963and current data: Survey of Current Business.

Data are compiled by the Portland CementAssociation in Chicago. They represent theyardage of concrete pavement in contractawards for roads, streets, and alleys in theUnited States (including Alaska beginningJanuary 1959, and Hawaii beginning Novem-ber 1960).

The annual figures include all areas reportedand all adjustments and corrections. For

1935—1963 they were computed from themonthly averages, shown separately for roadsand for streets and alleys, in Business Statistics.For years prior to 1935, they are the annualtotals reported by the Association. (See descrip-tive note, p. 156, Survey of Current Business,1936 Supplement.)

Monthly data covering 4- or 5-week periodsare also available from the same source.Irregular reporting for some areas before 1959,and adjustments and corrections which arenot allocated monthly, account for the dis-crepancies between the monthly and annualseries.

Concrete pavement data are shown sepa-rately in the source for (1) airports; (2) roads,and (3) streets and alleys.

Prior to 1933, airports were included withstreets and alleys, but the volume for airportswas negligible.

B. Building Permits and Housing StartsLONG'S INDEXES OF BUILDING PERMIT ACTIVITY, ANNUAL AND MONTHLY

General Description

SOURCE: Clarence D. Long, Jr., BuildingCycles and the Theory of Investment, Prince-ton University Press, 1940.

Information on number and value of build-ings covered by building permits issued orbuilding plans filed was compiled by Longfor individual cities from published reportsof city building departments or other munici-pal agencies, and from data collected(beginning in 1920) by the United StatesBureau of Labor Statistics. Supplementarysources of data were board of trade and cham-ber of commerce year books; and periodicals,including Bradstreet's, American Contractor,Construction News, Chicago Economist, RealEstate Record and Builders' Guide and Amer-ican Land and Title Register.

From these detailed data, several compositeindexes of building activity were constructed:(1) a monthly index of the value of all build-ing, (2) a set of annual indexes of the numberof new buildings covered by permits, by typeof building, and (3) a set of annual indexesof the value of total building, by type of build-ing. For specific information on coverage andmethod of computation, see individual seriesdescriptions below.

In general, the indexes are based on anincreasing number of cities, beginning in eachcase with one city. The index for each monthor year was calculated by dividing the volumeof building in all cities covered for that periodby the average monthly (or annual) volumefor the same cities during the base period(1930 for the monthly index and 1920—1930for the annual indexes).

The value indexes are based on value ofbuilding as represented by builders' estimatesof cost at the time of application for permits.No attempt is made to adjust the data forchanges in construction costs. The indexes ofnumber represent number of buildings Cov-ered by permits or plans filed. This is not pre-cisely the same as number of permits, since inseveral large cities, particularly in the case ofresidential building, a single permit or planfiled may represent many buildings.

Excluded are some state and local publicbuilding, most federal building, and all ruralbuilding. The data are restricted to largercities of the U.S. They are dominated by theeast, particularly by Manhattan in the earlyyears.

In addition to the series selected for inclu-sion here, indexes are available in the sourceon value of detached dwellings, multifamilydwellings, private nonresidential building, andalterations, for fourteen cities, 1868—1935;value of alterations and total building fortwenty-seven cities, 1868—1935; and numberof families accommodated in sixteen localities,1871—1935. See also Long's Index of Value,Total New Building, Adjusted by Colean andNewcomb (Series B9).

For further discussion and evaluation of theindexes, see Long, Building Cycles; U.S.Bureau of the Census, Historical Statistics ofthe United States, 1789—1945, pp. 166 and 325;Miles L. Colean and Robinson Newcomb,Stabilizing Construction: The Record andPotential, New York, 1952, Appendix N; andMoses Abramovitz, Evidences of Long Swingsin Aggregate Construction Since the Civil War,New York, NBER, 1964, Appendix A, PartIII.

BUILDING PERMITS AND HOUSING STARTS 241

Individual Series

Bi. Index of Value, Total Building, Monthly,5 to 35 Cities, 1882—1917 (1930 = 100);Annual, 1 to 37 Cities, 1868—1939 (1930 =100).SOURCE, Monthly data: Long, Building

Cycles, Pp. 214—219; Annual data: U.S. Bureauof the Census, Historical Statistics of theUnited States, 1789—1945, Series H77.

The published monthly series covers theperiod 1868 through April 1940. The indexwas smoothed, and random and seasonal fluc-tuations eliminated, by means of the Macau-lay 43-term graduation. (See source, pp. 101and 223.) The original monthly indexes forthe segment 1882 through 1917 are presentedhere with seasonal adjustment by NBER. For1868—1874, coverage consisted of Manhattanonly. The number of cities included reachedfive in 1882, thirty-five in 1906, and thirty-seven in 1930. The specific cities covered arelisted in the source. The monthly index coversall types of buildings for which permits wereissued or plans filed and includes alterationsand repairs. The base is the monthly averagein 1930 of the aggregate value of building inthe cities included in the index for the par-ticular month being measured.

For a monthly series on value of buildingplans in Manhattan, beginning 1868, see SeriesBl8. For monthly data on value of buildingpermits for 120 cities, beginning 1911, com-piled by see Series Bl9.

The annual figures for this series were cal-culated in Historical Statistics by averagingLong's monthly indexes.

B2. Index of Value, Total New Building, 1to 27 Cities, Annual, 1868—1935 (1920—30 = 100).SOURCE: Long, Building Cycles, Appendix

B, Section 2.

This series and Series B3 and B4 are basedon the same list of cities. The coverage con-

sisted of Manhattan only through 1874, andwas increased gradually thereafter until thefull complement of twenty-seven cities wasincluded in 1911. Value of alterations in thesecities is shown in the source as a separate series.The base is the annual average in 1920—1930of the aggregate value of building, in therespective categories, in the cities includedin the index for the particular year beingmeasured. See also Long's Index of Value,Total New Building, Adjusted by Colean andNewcomb (Series B9).

B3. Index of Value, Total Residential Build-ing, 1 to 27 Cities, Annual, 1868—1936(1920—3 0 = 100).SOURCE: See source note to Series B2.

B4. Index of Value, Total NonresidentialBuilding, 1 to 27 Cities, Annual, 1868—1936 (1920—30 = 100).SOURCE: See source note to Series B2.

B5. Index of Number, Total New Buildings,I to 29 Cities, Annual, 1856—1935 (1920—30 = 100).SOURCE: Long, Building Cycles, Appendix

B, Section 3.

This series and Series B6 and B7 are basedon the same list of cities. The coverage waslimited to Philadelphia through 1862. In 1863Manhattan and the Bronx were added. Thenumber of cities included was increasedgradually until the full complement of twenty-nine cities was reached in 1912. The base isthe annual average in 1920—1930 of the totalnumber of buildings, in the respective cate-gories, in the cities included in the index forthe particular year being measured.

B6. Index of Number, Total ResidentialBuildings, 1 to 29 Cities, Annual, 1856—1936 (1920—30 = 100).SOURCE: See source note to Series B5.


B7. Index of Number, Total NonresidentialBuildings, 1 to 29 Cities, Annual, 1856—1936 (1920—30 = 100).SOURCE: See source note to Series B5.

B8. Index of Value, Public Building, 1 to 11Cities, Annual, 1868—1935 (1920—30 =100).SOURCE: Long, Building Cycles, Appendix

B, Section 1.

This series is based on a list of cities forwhich buildings could be classified as detacheddwellings, multifamily dwellings, public build-ing, or private nonresidential. Value of altera-tions for all classes combined is shown in thesource as a separate series. Coverage waslimited to Manhattan through 1874, and wasincreased gradually until it reached fourteencities in 1911. The base is the annual averagein 1920—30 of the aggregate value of publicbuilding in the cities included in the indexfor the particular year being measured.

B9. Index of Value, Total New Building,Adjusted by Colean and Newcomb,Annual, 1868—1935 (1920—1929 = 100).

SOURCE: Miles L. Colean and RobinsonNewcomb, Stabilizing Construction: TheRecord and Potential, New York, 1952, Appen-dix N, Table 2.

This series represents an adjustment ofLong's index of permit values for new build-ing in one to twenty-seven cities (see SeriesB2). Two modifications were made. First, theresidential building component was given aweight of one-third, rather than the implicitweight of one-half in Long's series. Second,the base was changed from 1920—1930 to 1920—1929.

The result of this adjustment was one stepin the development by Colean and Newcombof an "Index of Dollar Volume of New Con-struction, 1831—1951," presented in the source,Appendix N, Table 3.

To calculate a constant price series, Abram-ovitz divided this index by Riggleman's indexof building costs. The resulting series for1853—1933, on a 1913 base, is shown in MosesAbramovitz, Evidences of Long Swings inAggregate Construction Since the Civil War,New York, NBER, 1964.

CONSTRUCTION AUTHORIZED BY BUILDING PERMITS, MONTHLY SERIESOF THE BUREAU OF LABOR STATISTICS AND THE BUREAU OF THE CENSUS

General Description

SOURCE: Estimates of monthly building permitactivity, including data on the value of totalbuilding and new nonresidential buildingand on the number and valuation of new pri-vate housing units authorized, have beenpublished by the Bureau of Labor Statisticsand the Bureau of the Census. The BLS data,from January 1942 to June 1954, covered allurban areas, with estimates for nonreportingor non-permit-issuing areas. The series onnumber of new dwelling units authorized inurban areas can be extended back from 1942to 1939 by the BLS series on "new family

dwelling units started." 1 For January 1954to April 1960, the BLS data covered 6,600 per-mit-issuing places, both urban and rural non-farm. In 1960 the BLS estimates were replacedby Census Bureau series covering 10,000 per-mit-issuing places. The series on number andvaluation of new private housing units werecarried back to January and continuedthrough 1963, and those on total building and

1 Data on dwelling units started were based onbuilding permits issued. Prior to 1945 no adjustmentwas made for lapsed permits or time lag betweenpermit issuance and start of construction, since it wasbelieved these influences were negligible at that time.(See Construction in the War Years, 1942—45, BLSBulletin 915, 1948, p. 28.)


new nonresidential building were carried backto June 1959. Beginning in January 1963,the coverage of the Census Bureau series wasagain expanded from 10,000 to approximately12,000 permit-issuing places.

The series presented here are currently avail-able in Construction Review, a monthly pub-lication of the Business and Defense ServicesAdministration. In addition, the series on newprivate residential construction authorized areavailable in Construction Reports, BuildingPermits, C42 reports (monthly) of the Bureauof the Census. Construction Reports, HousingStarts, C20 reports of the Census Bureau,include unadjusted and adjusted monthly dataon number of new private housing unitsauthorized. (See below for specific sources ofindividual series presented here.)

These series represent volume of construc-tion authorized, rather than actual construc-tion activity. (For a description of the use ofresidential building permit data in developingstatistics on housing starts and on value of newconstruction put in place, see Series B14 andC27.)

The total building series for the urban and6,600-place segments is composed of (1) newnonresidential buildings, including commer-cial and industrial buildings; educational,institutional and religious buildings; publicbuildings; public utility buildings; privateresidential garages; and an "all other" cate-gory; (2) private new residential housekeepingunits, excluding conversions and other itemsunder (4) and (5); (3) publicly financed resi-dential housekeeping units; (4) nonhousekeep-ing residential buildings, such as transienthotels, motels, nursing homes, and dormi-tories; and (5) additions, alterations, andrepairs to both residential and nonresidentialbuildings. Excluded are demolition or movingof buildings; nonbuilding construction, e.g.,streets, pipe lines; costs of land and land devel-opment; and architectural and engineeringfees. Although building permits generallyapply only to private building, the series on

total and nonresidential buildings authorizedinclude estimates of public construction inthese segments. Data were collected from build-ing permit officials through a BLS question-naire. Officials were asked to report on struc-tures covered by building permits, and onstate and local government building construc-tion in their localities for which permits werenot required. Information on constructioncontracts awarded for federal projects wasobtained from federal agencies. The 10,000-place, and current 12,000-place, segments forthe series on total and new nonresidentialbuilding authorized by building permits coverprivate construction only.

The current series, beginning 1963, repre-sent building authorized in all of the approxi-mately 12,000 permit-issuing places, so identi-fied in 1962, in the U.S. (including Alaska andHawaii). The monthly figures are compiledby direct tabulation of reports from approxi-mately 3,500 of the more active permit-issuingplaces, to which are added estimates for theremaining 8,500 places, based on reports froma representative sample of 500. It is estimatedthat for private residential construction, thepermit-issuing universe of 12,000 areasaccounted for about 83 per cent of all suchconstruction in the U.S. in 1963.2 The 1959to 1963 segment for 10,000 places was compiledin a similar manner, based on the permit-issuing places identified in 1959, with directtabulation of reports from 3,014 places, andestimates for the remaining places based on asample of 500. Annual totals for 10,000 placesin 1960—63, and for 12,000 places in 1963, werebased on a complete enumeration of all per-mit-issuing places identified. The BLS seriesfor January 1954 to April 1960 was based ona monthly survey of all of the 6,600 placesidentified as having permit systems in 1954.The earlier BLS series on number of residen-tial units, January 1939 to June 1954, andvalue of residential units, total and nonresi-dential building, January 1942, to June 1954,

2 Construction Reports, Building Permits, Bureau ofthe Census, No. December 1964, p. 3.


covered all places defined as urban in the1940 Census, and included estimates for someof the smaller urban places which did nothave permit systems. The 1954—60 segmentexcludes estimates for the non-permit-issuingplaces; however, coverage was extended due tothe inclusion of (1) areas which increased insize after the 1940 Census to become classifiedas urban, (2) areas incorporated after the 1940Census, and (3) unincorporated permit-issuingplaces, such as towns. (For a more comprehen-sive comparison of coverage between the 1942—54 urban building series and the 1954—60series for the permit-issuing universe, andcomments on linking the segments, see Trendsin Building Permit Activity, BLS Bulletin1243, 1959, pp. 4, 5, and 12.)

The method of reporting valuation on build-ing permits leads to serious limitations in thenationwide figures on dollar volume of build-ing permits. The permit value figures do notrepresent actual construction costs. They arereported differently in different permit-issuingplaces. In some cases they are estimated at aflat value per square foot, the value frequentlyhaving been fixed several years before; inother places the official may accept the figuresubmitted by the permit applicant. The totalvalue data are therefore less meaningful thanthe corresponding data on number of units,and should be used only with these limita-tions in mind.

Data are not adjusted for lapses of permitsor time elapsed between issuance of permitsand actual start of construction.

More detailed descriptions of the buildingpermit activity series can be found in theBureau of the Census publications, Construc-tion Reports, Building Permits, HousingAuthorized in Permit-Issuing Places: SummaryStatistics 1963, No. C42—53, 1964; ConstructionReports, Building Permits, New ResidentialConstruction Authorized in Permit-IssuingPlaces, Nos. C42—1 (September 1960) andC49—l (September 1960); and in the Bureauof Labor Statistics publication, Trends inBuilding Permit Activity, Bulletin 1243, 1959.

Two monthly bulletins from the Bureau of

the Census make available data on new hous-ing units authorized in specific areas. Forgeographic regions, states, and selected stand-ard metropolitan statistical areas, data onnumber and value of units by type of structureare included in the C42 reports. Data on num-ber of new housing units authorized in selectedpermit-issuing places are shown in the C40reports.

Individual Series

BlO. Total Building Authorized by BuildingPermits, Value, 1942—1963.

SOURCE:January 1942—June 1954 (urban building):U.S. Bureau of Labor Statistics, special tabula-tion, "Urban Building Authorized, by Prin-cipal Class of Construction and Type of Build-ing." January 1954—April 1960 (6,600 permit-issuing places): 1954—1956, Trends in BuildingPermit Activity, BLS Bulletin 1243, 1959.1957—April 1960, Construction Review, issuesof April 1958 to August 1960. June 1959—December 1963 (10,000 and 12,000 permit-issuing places): Construction Review, issuesof September 1960 to July 1965.

This series and Series B!! were seasonallyadjusted by NBER.

B!!. New Nonresidential Building Authorizedby Building Permits, Value, 1942—1963.

SOURCE: Same as Series BlO.

B!2. Number of New Private Housing UnitsAuthorized by Building Permits, 1939—1963.

SOURCE:January 1939—June 1954 (urban places):1939—1941, Construction in the War Years,1942—45, BLS Bulletin 915, 1948, Table 15.1942—1954, U.S. Bureau of Labor Statistics,special tabulation, "Urban Building Author-ized, by Principal Class of Construction andType of Building." January 1954—April 1960(6,600 permit-issuing places): 1954-1956,Trends in Building Permit Activity, BLS


Bulletin 1243, 1959, Tables B-5 and B-6.1957—1 960, Construction Review, issues ofApril 1958 to August 1960. January 1959—September 1964 (10,000 and 12,000 permit-issuing places): Construction Reports, Hous-ing Starts, No. C20—50, 'August 1963, Table6; No. C20—63, September 1964, Table 6.

Seasonal adjustment for the urban and6,600 permit places segments was by theBureau of the Census for NBER. For the10,000 and 12,000 places segments, the adjust-ment was by the source.

B13. Value of New Private Housing UnitsAuthorized by Building Permits, 1942—1963.

SOURCE:January 1942—June 1954 (urban places): U.S.Bureau of Labor Statistics, special tabulation,

"Urban Building Authorized, by PrincipalClass of Construction and Type of Building."January 1954—April 1960 (6,600 permit-issuingplaces): 1954—1956 (annual data), Trends inBuilding Permit Activity, BLS Bulletin 1243,1959, Table B-7. 1954—1956 (monthly data),Construction Review, issues of April 1955 toApril 1957. 1957—1960, Construction Review,issues of April 1958 to August 1960. January1959—December 1963 (10,000 and 12,000 per-mit-issuing places): 1959 (annual data), Con-struction Reports, Building Permits, No.C42—43, October 1963. 1959 (monthly data)ibid., No. C49—1, September 1960. 1960—1963(annual data), ibid., No. C42—53. 1960—1963(monthly data), ibid., Nos. C42—9 to C42—60.

Seasonal adjustment by Bureau of the Cen-sus for NBER.

HOUSING STARTS, BLANK-BLS-CENSUS, ANNUAL AND MONTHLY

General Description

SOURCE: The four monthly series on privatenonfarm housing starts presented here repre-sent one major component of the total hous-ing starts series currently compiled by theBureau of the Census, and until 1959 pre-pared by the Bureau of Labor Statistics. Theaggregate series includes public as well asprivate housing, and beginning 1959, farm aswell as nonfarm construction. The series areavailable currently in Construction Reports,Housing Starts, C20 reports, of the Bureauof the Census. A breakdown of the aggregateinto metropolitan and nonmetropolitan loca-tion is available in this source.

"A housing start is the start of constructionon a new housing unit, when located withina new building which is intended primarilyas a housekeping residential building designedfor non-transient occupancy. Start of construc-tion for private housing units is defined asthe beginning of excavation for the foundation

of the building. All housing units in a multi-family building are counted as being startedwhen excavation for the building is started.A housing unit is defined as a single room orgroup of rooms intended for occupancy asseparate living quarters by a family, by agroup of unrelated persons living together,or by a person living alone." 'The series thusexclude group quarters (such as dormitoriesand rooming houses), transient accommoda-tions, family units in primarily nonresidentialbuildings, units in structures which are moved,units provided by conversion of residentialor nonresidential space, and mobile homes(trailers).

Private housing units include those financedby a government insured or guaranteed mort-gage. One-family structures include detached,semidetached, or row houses, which haveseparate entrances, heating, and utility connec-tions, and which can be sold independently.Two-family structures have one unit over the

1 Construction Reports, Housing Starts, No. C20—60,Bureau of the Census, June 1964, p. 5.


other, or two on the same floor, with acommon entrance or common heating facil-ities. Three or more family structures containthree or more units having a common base-ment, heating plant, stairs, or entrance.

The current monthly series on private non-farm housing starts was initiated by theBureau of the Census in May 1960, and at thattime was carried back on a modified basis toJanuary 1959. This series replaced the Bureauof Labor Statistics series, which covered theperiod January 1939 through April 1960. Thedata are derived by adding separate estimatesfor housing starts in localities requiring andnot requiring building permits. Starts in per-mit areas are based on building permit data(see Series B12). Estimates of the number ofhousing units authorized in all permit-issuingplaces are converted to starts estimates byadjusting them for lapses of permits and timelag between permit issuance and start of con-struction. Information from a continuingsample field survey is the basis for these adjust-ments. An additional adjustment is then madeto account for housing construction startedin advance of (or without) the required per-mit authorization. Estimates of housing startsin nonpermit areas are developed from a con-tinuing monthly survey, conducted in a sam-ple of fifty-six large areas. Data are gatheredfrom informed sources including public offi-cials, builders, suppliers, lending institutions,and public utility companies. Reports fromthese sources are verified by field visits. Anadjustment for underreporting by thesesources is then made on the basis of an inten-sive canvass by visual inspection of a subsam-ple of land area. A detailed description ofthese procedures is included in ConstructionReports, Housing Starts, Bulletin C20—11(Supplement), Bureau of the Census, May1960.

The new series on housing starts, describedabove, represents a substantial improvementover the old series in coverage and in measure-ment of month-to-month changes, and the dataare therefore not comparable to the old series.

In May 1964, the Bureau of the Census releasedrevised annual estimates for 1945—58, whichare comparable to the current monthly series.These revisions are based on an analysis ofthe 1950 and 1960 Censuses of Housing, the1956 National Housing Inventory, and the1959 Survey of Components of Change andResidential Financing (all compiled and pub-lished by the Bureau of the Census). A detaileddescription of these revisions is included inConstruction Reports, Housing Starts, BulletinC20—60, Bureau of the Census, June 1964.

The revisions for the period 1945—58 werenot developed on a monthly basis by theBureau of the Census. However, the Businessand Defense Services Administration has pre-pared monthly housing starts figures for thisperiod, which conform to the level of therevised annual series, and which follow essen-tially the pattern of month-to-month changesof the former BLS series. These monthlyestimates are shown here for total private non-farm housing starts (Series B14). For an expla-nation of the method used in deriving themonthly estimates, see Construction Review,July 1964.

Estimates at the revised level are not avail-able for this period for the component serieson one-, two-, and three-or-more-family units;therefore, the figures for these three series donot add to the aggregates for the period 1945—58. According to the Bureau of the Census,"It seems probable that the major part, orperhaps even substantially all, of the revisionsis applicable to one-family structures, andthat little change is implied in the figures forstructures containing three-or-more families." 2

The former monthly BLS series, shownhere for 1939—45 (total), and 1940—59 (bytype of dwelling), were based on buildingpermit data, and field surveys in nonpermitareas. They differed from the current Censusseries in that they were based on a smallernumber of permit-issuing places; they did notmeasure directly, on a continuing basis, con-

2 Construction Reports, Housing Starts, report C20—60, June 1964.


version of permits to starts, and undercover-age in permit-issuing places; and they did notinclude as intensive a survey of nonpermitplaces. For detailed descriptions of the esti-mating procedures used in these series, includ-ing major revisions in 1941—42, 1946—47, andin 1954, see Nonfarm Housing Starts 1889—1958, BLS Bulletin 1260 (1959), pp. 1—11;Techniques of Preparing Major BLS Statisti-cal Series, BLS Bulletins 1168 (1954), Chapter2, and 993 (1950), Chapter 3.

The earlier annual data, based on theinvestigations of David M. Blank, and ofDavid L. Wickens and Ray R. Foster, were

developed from building permit data for

urban areas, expanded to approximate total

private nonfarm housing starts by relating

them to changes in population.

For 1889—1919, the BLS adopted the seriescalculated in David M. Blank, The Volumeof Residential Construction, 1 889—1950, NewYork, NBER, 1954. The basic source forBlank's estimates was an extensive set ofbuilding permit data transcribed from local

records in 1938—40 as part of a WPA projectsponsored by the BLS. After stratifying the

reporting cities by size, Blank estimated the

number of starts in each city size class and inall cities on the basis of data on populationchange in permit-issuing and non-permit-issuing places, and then calculated total non-farm housing starts by using his urban esti-mates and data on urban and rural nonfarmpopulation changes. Since the series are

annual, no adjustment was made for the lagbetween permit issuance and the start of con-struction. Only minor allowances for lapsesof permits, for multifamily dwellings in NewYork City, were calculated in this part ofBlank's series.

The BLS series for 1920—29 incorporatesthe estimates of Wickens and Foster, whichare based on somewhat similar methods. The1930—36 data are BLS revisions of the Wick-ens-Foster estimates. The 1937—39 data areBLS estimates, based on building permit data.For detailed descriptions of the annual data,

see Blank, The Volume of Residential Con-struction, 1889—1950, David L. Wickens andRay R. Foster, Non farm Residential Construc-tion, 1920—1936, New York, NBER, 1937;David L. Wickens, Residential Real Estate,New York, NBER, 1941; and BLS Bulletinscited above.

There are also available, in Blank's TheVolume of Residential Construction, esti-mates of number and permit value of newprivate urban housekeeping units started,1890—1929 (Table 10, p. 41).

The Blank data back to 1889 cover sub-stantially more cities than the Long series(see Series Bl through B9). Coverage wasgradually increased from 1 city in 1870 to 20cities in 1889, 104 cities in 1906, 156 in 1911,255 in 1918, and 314 in 1929.

For a regional breakdown of the estimatesof new private nonfarm housing units startedin 1920—50, see Leo Grebler, David M. Blank,and Louis Winnick, Capital Formation inResidential Real Estate, Princeton UniversityPress for NBER, 1956, Table H-i, p. 396.

Individual Series

B14. Number of New Private Non farm Hous-ing Units Started, Annual, 1889—1963;Monthly, 1939—1963.

SOURCE:1889—1919: David M. Blank, The Volume ofResidential Construction, 1889—1950, NewYork, NBER, Technical Paper 9, 1954. 1920—1938: Non farm Housing Starts 1889—1958, BLSBulletin 1260, 1959. 1939—1945 (old series):Construction During Five Decades, HistoricalStatistics, 1907—52, BLS Bulletin 1146, 1954.1945—1958 (new series, annual data): Con-struction Reports, Housing Starts, No. C20—60, June 1964. For monthly data, ConstructionReview, July 1964. 1959—1963: ConstructionReports, Housing Starts, Nos. C20—50 to C20—66, August 1963 to December 1964.

Seasonally adjusted by compiling agencies,except figures for 1945 (new series) which


were seasonally adjusted by NBER by apply-ing implicit seasonal factors from the BLSseries for 1945.

B15. Number of New Private Non farm Hous-ing Units Started, One-Family, Annual,1900—1963; Monthly, 1910—1963.

SOURCE:

1900—1919: Blank, Volume of Residential Con-struction. 1920—1939: Non farm Housing Starts1889—1958, BLS Bulletin 1260, 1959. 1940—1959 (old series): Housing Statistics, Histori-cal Supplement, June 1960, Housing andHome Finance Agency. 1959—1963 (new series):Construction Reports, Housing Starts, Nos.

C20—50 to C20—66, August 1963 to December1964.

The series was seasonally adjusted byNBER.

B16. Number of New Private Non farm Hous-ing Units Started, Two-Family, Annual,1900—1963; Monthly, 1940—1963.

SOURCE: See source note to Series Bl5.

B17. Number of New Private Non farm Hous-ing Units Started, andOver, Annual, 1900—1963; Monthly,1940—1963.

SOURCE: See source note to Series B15.

OTHER SERIES ON BUILDING PERMITS, ANNUAL AND MONTHLY

B18. Value of Plans for New Buildings, Man-hattan, Monthly, 1868—1910.

SOURCE:

New York City Record (official daily journalof New York City Record Office); Real EstateRecord and Builders' Guide (F. W. DodgeCorp.); and Annual Reports, Building Depart-ment, Borough of Manhattan, and Depart-ment of Housing and Buildings, New YorkCity.

The data represent estimated cost of build-ings as indicated on building plans filed. Noadjustment is made for plans later withdrawnor disapproved. Only new buildings areincluded in this series; data on additions andalterations are available from the samesources. Types of building included are resi-dential, commercial, industrial, municipal,educational, religious, and institutional.

The 1868—1897 figures include Manhattanand such portions of the Bronx as wereannexed from time to time. For January1898 through January 1899, Manhattan andthe entire Bronx are included since data forthe two boroughs were not separable. Begin-ning February 1899, separate data are avail-able on Manhattan, and the last segment,February 1899 to December 1940, covers Man-

hattan only. An overlap for February 1899to December 1900 is provided, showing Man-hattan and Bronx combined. For certain years,weekly figures were converted to monthly byNBER. In these cases, data for overlappingweeks were prorated according to the numberof days in each month.

The building plans data differ from build-ing permit data in that they include value ofbuilding plans for which permits are neverissued, and therefore cover some proposedconstruction that does not materialize. Thedata have been used, however, in compilationsof building permit data. (See Series B20, andSeries Bi, both of which cover additions andalterations, as well as new construction.)

Later annual data can be found in annualreports of the Department of Housing andBuildings (later, Department of Buildings).

The series was seasonally adjusted byNBER.

B19. Value of Building Permits, Monthly, 20Cities, Babson, 1903—22; 120 Cities, Dunand Bradstreet, 1911—September 1960.

SOURCE:1903—1922: Babson's Statistical Organization.191 1—February 1955: Dun and Bradstreet,


Monthly Review, September 1935; Dun'sStatistical Review, January 1939, May 1947,November 1953, and April 1955. MarchSeptember 1960: Dun and Bradstreet, unpub-lished data.

Building permit data from two sourceshave been combined in this series. The firstsegment, for the period 1903 to 1922, wascompiled by the Babson's Statistical Organi-zation (now Babson's Reports), and coverstwenty large cities which accounted for ap-proximately one-third of U.S. urban popula-tion in 1920. These data were obtained di-rectly from Babson's by the NBER. They werealso published (and extended through Novem-ber 1926) in "Babson's Desk Sheet of Tables,"1911 to 1926.

For the period January 1911 to September1960, data were compiled for 120 identicalcities by Dun and Bradstreet (formerly Brad-street's). Data for March 1955 and later monthswere not published. Reports were furnishedto the compilers by the building departmentsof the individual cities. The 120 cities includeall the twenty cities in the Babson series, andaccounted for slightly over half the U.S. urbanpopulation in 1920. The figures representtotal costs estimated by builders, under per.mits issued within the corporate limits of thecities covered. They include new building,additions, alterations, and repairs. After Sep.tember 1960 only quarterly data are availablefor this series.

Both segments of this series were seasonallyadjusted by NBER. They have been linked atthe year 1911 for cyclical analysis purposes(see Section IV).

A monthly index of building activity forthe period February 1903 to December 1933,developed by William H. Newman, alsoemploys these two series: Babson's for 1903—1910, and Bradstreet's for 1911—1933. (SeeWilliam H. Newman, "The Building Indus-try and Business Cycles," The Journal ofBusiness of the University of Chicago, July1935, pp. 63—71.

Riggleman and Riggleman-IsardSeries on Value of Building Permits

B20. Value of Building Permits Per Capita,Current Dollars, Riggleman, Annual,1 830—1933.

SOURCE: John R. Riggleman, "Variationsin Building Activity in United States Cities,"unpublished Ph.D. dissertation, The JohnsHopkins University, 1934.

The basic data for this series and for SeriesB21 and B22 are Riggleman's compilationsand estimates of the value of building permitsfor an increasing number of U.S. cities. Inthe early years they are largely estimates,based mainly on other information, such asreal estate valuations. Value figures are forresidential and nonresidential building com-bined, including additions and alterations.The data cover Manhattan, Boston, and Wash-ington, D.C., from 1830 through 1848. Thecoverage then increases gradually to a totalof ten cities by 1868, and to sixty-five citiesfor 1900 and subsequent years. (For specificcities covered each year, see Colean and New-comb, Stabilizing Construction, Appendix N,Table 4, fn. c, p. 240.) The series are domi-nated by Manhattan before 1870, and to someextent for a few years after that.

This series represents the aggregate valueof building permits in the cities covered,divided by a combined population estimatefor these cities. The population estimates arebased on Census figures, with straight-lineinterpolation between Census years.B21. Value of Building Permits Per Capita,

Constant Dollars, As Per Cent of Trend,Riggleman, Annual, 1830—1 933.

SOURCE: See source note to Series B20.

This series represents the per capita figuresin Series B20, deflated by a cost index (1913 =100), and expressed as percentages of the ordi-nates of a long-term trend line. The costindex was constructed by Riggleman usingAmerican Appraisal Co. indexes, EngineeringNews-Record indexes, and, for earlier years,a combination of building wage rates and


softwood lumber prices. See Colean and New-comb, Stabilizing Construction, Appendix N,Table 4, fn. b, p. 240.

B22. Index of Value of Building Permits,Riggleman Adjusted by Isard, Annual,1 83 0—1 933.

SOURCE: Miles L. Colean and RobinsonNewcomb, Stabilizing Construction: The Rec-ord and Potential, New York, 1952, AppendixN, Table 2. See also source note to SeriesB20.

Riggleman's basic data were adjusted byWalter Isard in "The Economic Dynamics

of Transport Technology" (an unpublishedPh.D. dissertation, Yale University, 1947). Theresult was an index similar in construction toLong's indexes (see Series Bi and B2). Thepopulation adjustment of Riggleman's seriesis eliminated. The permit values in an individ-ual city are converted to relatives of theaverage annual value of permits in the samecity during the period 1920—29. The combinedindex for each year is then calculated as theweighted average of the relatives for thecities covered in that year, with weights pro-portional to permit values in these cities inthe base period.

C. Construction Expenditures and Related Measures

ESTIMATES OF GROSS AND NET CONSTRUCTION, KUZNETS, ANNUAL

General Description

SOURCE: These series are a selection fromthe large number of annual estimates, partlyunpublished, which underly the five-year mov-ing averages published with explanatory notesin Simon Kuznets, Capital in the AmericanEconomy: Its Formation and Financing,Princeton University Press for NBER, 1961.

Many of the series incorporate BLS-Commerce estimates of construction put inplace, which have been superseded, exceptfor comparison with earlier years, by therevised figures given later in this volume andin the official publications.

Because the Kuznets series were presentedonly in the form of five-year averages, somecrude extrapolations and interpolations couldbe used in constructing the series which arenot suitable for use as annual series. Theseseries, or parts of series, are not shown here,but some are noted in the individual seriesdescriptions. The page and table numbers inthe source notes to individual series all referto Capital in the American Economy unlessotherwise specified.

Individual Series

Cl. Gross Total Construction, Including Oiland Gas Well Drilling, Current Dollars,1889—1955.

C2. Gross Total Construction, Including Oiland Gas Well Drilling, 1929 Dollars,1889—1955.SOURCE: Annual estimates underlying

five-year moving averages from Kuznets, TableR-30, pp. 576—586.

For years since 1915 these are mainly BLS-Commerce estimates. Therefore, aside fromthe 1915—28 figures for oil and gas well drill-ing, Series Cl is superseded, except for com-parison with earlier years, by Series C65.

Kuznets' current price figures for 1952—55are the Commerce Department estimates of"Total New Construction Activity" plus"Petroleum and Natural Gas Well Drilling"from the Survey of Current Business, July1956, Table 31, p. 22. The constant (1929)price estimates for 1953—55 were calculated,for private construction, from the current priceestimates and an implicit price index derivedfrom current and constant dollar figures inthe Economic Report of the President, January1956, pp. 165 and 168, and for public con-struction, by extrapolating the 1952 estimateby the constant price series in ConstructionReview, October 1956, p. 18.

For earlier years, Kuznets calculated totalgross construction as the sum of (1) cost ofoil and gas wells drilled (see Series C84 andC85) and (2) all other construction.

The value of "all other construction" in1929 dollars was calculated as follows: 1889—1914: Output of construction materials fordomestic consumption, multiplied by 1.54056,the ratio of new construction to cost of con-struction materials consumed in 1919—33. Themethod of estimation is described in Kuznets,Capital in the American Economy, TableR-30, notes to col. 10, and Kuznets, NationalProduct since 1869, Table 11-5, notes tocol. 7. For data on output of constructionmaterials, see William H. Shaw, Value ofCommodity Output Since 1869, New York,NBER, 1947 (Series D7). 1915—1920: Sumof (1) new private nonfarm residential con-struction including additions and alterations,


in Grebler, Blank, and Winnick, Capital For-mation in Residential Real Estate, AppendixB, Table B-6, and (2) other new constructionfrom Construction and Building Materials,Statistical Supplement, May 1954, Tables 15and 16, adjusted to 1929 prices. 1921—1952:Given in Construction and Building Mate-rials, Statistical Supplement, May 1954, Tables15 and 16, adjusted to 1929 prices.

The calculations for current prices wereas follows: 1889—1918: Calculated by multi-plying the series in 1929 prices by the under-lying price index. The latter was computedfor 1915 and later years from Constructionand Building Ma teria is, Statistical Sup pie-ment, May 1954, pp. 33 and 82, and was ex-trapolated from 1915 by the total construc-tion cost index described in Kuznets NationalProduct since 1869, Table IV-4, notes toline 1. 1919—1951: Sum of (a) new privatenonfarm residential construction includingadditions and alterations and excluding non-housekeeping, from Grebler, Blank, and Win-nick, Capital Formation in Residential RealEstate, Appendix B, Table B-6, and (b) allother new construction, from Constructionand Building Materials, Statistical Supple-ment, May 1954, Tables 2 and 3.

C3. Gross Private Non farm Residential Con-struction, Current Dollars, 1889—1955.

C4. Gross Private Non farm Residential Con-struction, 1929 Dollars, 1 889—1955.SOURCE: Kuznets, Table R-30, pp. 576—

587. See general description.

Gross private nonf arm residential construc-tion is composed of housekeeping residentialconstruction plus additions and alterations.

Current dollar figures were taken from thefollowing sources: 1889—1920: New construc-tion, from David M. Blank, The Volume ofResidential Construction 1889—1950, Table18, plus a rough estimate of additions andalterations from Leo Grebler, David M.

tion in Residential Real Estate, Table D-2,p. 363. The new construction series was builtup originally from the BLS-WPA buildingpermit surveys (see Blank, The Volume ofResidential Construction, p. 33). Blank esti-mated the permit value of construction startedand the total cost of construction started (seenotes to Series B14). He then converted con-struction started to construction put in place,by using an estimate of the average carry-over into the following year of constructionstarted in a given year (ibid., pp. 52—56).1921—1952: BLS-Commerce estimates pub-lished in Construction and Building Mate-rials, Statistical Supplement, May 1952, p. 6,and Survey of Current Business, July 1956,Table 31. They have now been superseded bymore recent Commerce Department estimates(Series C29).

For an allocation of nonfarm residentialconstruction expenditures, 1837 to 1914, bytype of structure, see Raymond W. Goldsmith,A Study of Saving in the United States, Prince-ton, 1955, Vol. I, pp. 616—619.

The constant (1929) dollar figures were cal-culated as follows: 1889—1955: The series incurrent prices was deflated by residential con-struction cost indexes. For 1889—1952, the in-dex in Grebler, Blank, and Winnick, CapitalFormation, Table B-10, Column 1, was used.This is the Boeckh residential constructioncost index extrapolated back from 1910 bybuilding-trade wage rates and building mate-rial prices. The index for 1953—55 was theindex implicit in the series for all privateresidential nonfarm building (including non-housekeeping) in 1947—49 prices, ConstructionReview, September 1956.

C5. Gross Private Construction, ExcludingNon farm Residential, Current Dollars,1889—1955.

C6. Gross Private Construction, ExcludingNon farm Residential, 1929 Dollars, 1889—1955.SOURCE: Kuznets, Table R-30, pp. 576—

Blank, and Louis Winnick, Capital Forma- 587. See general description.

CONSTRUCTION EXPENDITURES 253

These series are calculated as a residual bysubtracting the following series from "TotalConstruction" (Series Cl and C2): (1) "GrossPrivate Residential" (Series CS and C4); (2)"Gross Public Construction ExcludingMilitary" (Series C7 and C8); (3) "GrossMilitary Construction." (An earlier version ofBLS-Commerce Series C55 and C56. See notesto "Net Military Construction," Series C17and Cl8.)

C7. Gross Public Construction, ExcludingMilitary, Current Dollars, 1889—1955.

C8. Gross Public Construction, ExcludingMilitary, 1929 Dollars, 1889—1 955.SOURCE: Kuznets, Table R-30, pp. 576—587.

See general description.

Current dollar figures were taken from thefollowing sources: 1889—1914: Sum of separateestimates for federal, state, and local gov-érnments. The federal government series arederived mainly from those published in His-torical Statistics of the United States, 1789—1945, U.S. Bureau of the Census, 1949 (seeSeries C7l; C72; C73; C74; and C75 below),and the state and local government series arederived mainly from the Census Bureau'spublications on government finances (seeSeries C19 below). A detailed description isgiven in Kuznets, notes to Table R-30, pp.580—584. 1915—1955: BLS-Commerce estimatespublished in Construction and BuildingMaterials Statistical Supplement, May 1954,Table 3 ("total public construction" minus"military facilities"), and Survey of CurrentBusiness, July 1956, Table 31. They are super-seded by revised estimates (see Series C49 andC51).

The constant (1929) dollar figures weretaken from the following sources: 1889—1914:The components of the current price serieswere converted to 1929 prices by separateprice indexes, each derived by extrapolatingthat implicit in the estimate for 1915 by thecost of construction index described in Na-

tional Product Since 1869, Table IV-4, notes toline 1. (See Kuznets, Table R-30 notes forfurther detail.) 1915—1955: BLS-Commerceestimates, from Construction and BuildingMaterials, Statistical Supplement, May 1954,Table 16, adjusted to a 1929 price base, andextrapolated from 1952 to 1955 by the seriesin 1947—49 prices given in Construction Re-view, September 1956, Table 4, p. ii.

These BLS-Commerce figures have beensuperseded by later estimates (see Series C50and C56).

C9. Net Total Construction, Including Oiland Gas Well Drilling, Current Dollars,1889—1955.

ClO. Net Total Construction, Including Oiland Gas Well Drilling, 1929 Dollars,1889—1 955.

SOURCE: Kuznets, Table R-31, pp. 592—595. See general description.

Net total construction was derived by add-ing the four components: nonfarm residential(Series Cli and C12), private nonresidential(Series Cl3 and Cl4), public, excluding mili-tary (Series C15 and C16), and military(Series Cl7 and C18).

Kuznets' measures of capital consumptionare intended to include not only physical de-terioration but also depletion, obsolescence,accidental losses or damage to capital, anddemolition. They are calculated on a replace-ment-cost basis.

For an adjusted version of the Kuznetsseries, extended back to 1837, see Goldsmith,A Study of Saving, Vol. I, Tables R-22, p. 611;R-26, p. 617; and R-27, p. 619. Appendix Bof Raymond W. Goldsmith, The NationalWealth of the United States in the PostwarPeriod (Princeton for NBER, 1962) gives, for1945—58, estimates of gross investment, depre-ciation, and net investment in many types ofconstruction.

Cli. Net Nonfarm Residential Construction,Current Dollars, 1889—1955.


C12. Net Nonfarm Residential Construction,1929 Dollars, 1889—1 955.


Net nonfarm residential construction isobtained by subtracting capital consumption,including depreciation and demolition, fromgross construction (Series C3 and C4). Thedepreciation and demolition figures for 1889—1953, published in Grebler, Blank, andWinnick, Capital Formation, Table E-2, pp.384—385, and explained in Appendix E ofthat book, were extended to 1955 by Kuznetsfollowing the same methods.

For rough estimates of the corporate shareof gross and net new nonfarm 1—4 family andmultifamily construction, see Goldsmith, AStudy of Saving, Vol. I, Tables R-4, p. 587;R-5, p. 588; R-7, p. 591, and R-8, p. 592.

C13. Net Private Construction, ExcludingNon farm Residential, Current Dollars,1889—1 955.

C14. Net Private Construction, ExcludingNonf arm Residential, 1929 Dollars,1 889—1955.

SOURCE: Kuznets, Table R-32, pp. 592—595.See general description.

Net private construction excluding nonfarmresidential was derived by subtracting depre-ciation and depletion from the gross series(Series C5 and C6).

The constant (1929) dollar figures were cal-culated as follows: 1889—1918: Annual depre-ciation and depletion figures were derived byinterpolation and extrapolation from decadeestimates calculated from wealth data in Kuz-nets, National Product Since 1869. For detailssee notes to Kuznets, Capital in the American

Economy, Table R-17, p. 531. 1919—1955: (1)The Commerce Department series for depre-ciation on an original-cost basis, published inNational Income, 1954 Edition, Table 4, andSurvey of Current Business, July 1956, Table4, were extended back to 1919 by Fabricant's

series (Solomon Fabricant, Capital Consump-tion and Adjustment, New York, NBER,1938) and deflated by Fabricant's price index,extended to 1955 by Kuznets. For furtherdescription see Kuznets, Capital in the Amer-ican Economy, Table R-8, p. 501. (2) Depletionwas estimated as 8 per cent of depreciation(see ibid.).

Current dollar figures were calculated asfollows: The depreciation series in 1929 priceswas multiplied by the appropriate price indexdescribed in Kuznets, Capital in the AmericanEconomy, Table R-31, col. 4.

C15. Net Public Construction, Excluding Mili-tary, Current Dollars, 1889—1955.

C16. Net Public Construction, ExcludingMilitary, 1929 Dollars, 1889—1955.


"Net Public Construction, Excluding Mil-itary," was derived by subtracting deprecia-tion from "Gross Public Construction, Exclud-ing Military" (Series C7 and C8). The depre-ciation figures were derived as the sum ofseparate estimates for sewer and highwayconstruction, on the one hand, and all othergovernment nonmilitary construction on theother. The life for sewers and highways wasassumed to be twenty-five years, and all othergovernment construction, fifty years.

Depreciation in current prices was derivedby multiplying the depreciation series in 1929prices by the price index calculated for thegiven type of construction, described in thenotes to Kuznets, Table R-30.

C17. Net Military Construction,Dollars, 1914—1955.

Current

C18. Net Military Construction, 1929 Dollars,

1914—1955.



Net military construction is derived by sub-tracting depreciation from gross military con-struction. Kuznets' gross military constructionseries is taken from BLS-Commerce figures(Series C55). However, the Kuznets series addsan extension back to 1914 and does not includethe latest revisions in the government series.Kuznets derived his 1914 figure by averagingthe 1914 and 1915 government fiscal-yearfigures for military and naval construction(published in Historical Statistics of the U.S.1789—1945, Series H29) to obtain a calendar-year figure. The BLS-Commerce figures Kuz-nets used were those published in Construc-tion Volume and Costs 1915—56, Table 3, p. 7.Depreciation was calculated assuming atwenty-year life.

The current price depreciation series wasderived by multiplying the series in 1929prices by the implicit price index for militaryconstruction (gross military construction incurrent prices divided by gross military con-struction in 1929 prices). See note to Kuznets,Table R-6, col. 5.

For data through 1958 see Goldsmith,National Wealth, Tables B-l69, B-170, B-171,and related tables.

C19. Gross State and Local Government Con-struction, Current Dollars, 1902—1915.

SOURCE: Kuznets, Table R-30, pp. 580—584.See general description.

These figures are the sum of estimates forhighways (Series C20) and other construc-tion. The latter are mainly extrapolations ofthe BLS-Commerce estimates for 1915—1918given in Construction and Building Materials,Statistical Supplement, May 1954, by capitaloutlay figures from Census Bureau reports ongovernment finances, particularly the censusesof Wealth, Debt, and Taxation for 1902 and1913, the Financial Statistics of States, and theFinancial Statistics of Cities.

Very rough further extrapolations back to1819, partly for the purpose of calculating

depreciation, are described in the notes toKuznets, Table R-30, but are not shown here.

For rough estimates of state and local gov-ernment new construction back to 1837, seeGoldsmith, A Study of Saving, Vol. I, TablesR-26, p. 617, and R-27, p. 619.

C20. Gross State and Local Got'ernment High-way Construction, Current Dollars,1902—1915.


These are the sum of separate estimates forstate governments, counties, and local gov-ernments, all based mainly on Census Bureaureports on government finances, particularlythe censuses of Wealth, Debt, and Taxation,for 1902 and 1913, the Financial Statistics ofStates, and the Financial Statistics of Cities.A considerable amount of interpolation wasrequired to derive the estimates for states,counties, and small cities.

This series is treated in the source as a con-tinuation of the BLS-Commerce series pub-lished in Construction and Building Materials,Statistical Supplement, May 1954, even thoughthere were some differences between the twofor 1915 and later years.

A series for highway expenditures of 145cities, based on the same sources, was con-structed by Harold Wolkind in Fluctuationsin Capital Outlays of Municipalities, U.S.Department of Commerce, Bureau of Foreignand Domestic Commerce, Economic Series No.10, 1941.

Raymond Goldsmith has also publishedestimates back to 1897 of the capital outlay(which includes equipment) of state govern-ments for total and highway construction, andof local governments for streets and highways.See Goldsmith, A Study of Saving, Vol. 1,

Tables G-6, p. 1053, and G-15, p. 1067; andThe National Wealth of the United Statesin the Postwar Period, Princeton for NBER,1962, Tables B-136, p. 364, and B-l37, p. 365.


VALUE OF NEW CONSTRUCTION PUT IN PLACE, IN CURRENT AND CONSTANTDOLLARS, BLS-COMMERCE DEPAR TMENT, ANNUAL AND MONTHLY

General Description

SOURCE: For 1946 and subsequent years, themonthly and annual estimates were preparedby the Bureau of the Census. The monthlyand annual estimates for the years 1939through 1945, and annual estimates for 1915through 1938, were prepared and publishedjointly by the Business and Defense ServicesAdministration of the Department of Com-merce and the Bureau of Labor Statistics.1The Census estimates for 1946 through 1959represent revisions of the former BLS-BDSAseries. For series in which the revisions resultin a break in comparability, an overlap forthe year 1946 is shown. The series presentedhere, unless otherwise noted, are from thefollowing sources:

1915—1944: Construction Volume and Costs,1915—1956, Statistical Supplement to Volume3 of Construction Review, Business and De-fense Services Administration and Bureau ofLabor Statistics, 1958. 1945—1946: Value ofNew Construction Put in Place (New Con-struction Expenditures), 1945—58, RevisedJune 1959, Business and Defense Services

1 Effective July 1, 1959, responsibility for compila-tion of the construction activity estimates was trans-ferred to the Bureau of the Census from the Bureauof Labor Statistics and the Business and Defense Serv-ices Administration. Prior to that time, the BLS wasresponsible for the estimates on new private nonfarmhousekeeping residential construction and all publicconstruction, while the BDSA was responsible for thoseon all other types of private construction. The seriesprepared by these two agencies were an extension ofestimates covering the years 1915 to 1937, developedoriginally in the late 1930's by the Bureau of Foreignand Domestic Commerce.

Since the preparation of this description, the Bureauof the Census has initiated some new procedures forderiving the estimates of new construction. (See Con-struction Reports, Construction Activity, C30-65S,January 1966; and Construction Statistics, 1915—1961,Supplement to Construction Review, BDSA, January1966.)

Administration and Bureau of Labor Statistics.1946—1963: Value of New Construction Put inPlace, 1946—1963 Revised, Construction Re-ports, Construction Activity, C30—61 Supple-ment, Bureau of the Census, October 1964.After the tables for this volume were com-pleted, new summaries, revisions, and descrip-tions of these data were published in Con-struction Statistics, 1915—1964, Supplement toConstruction Review, January 1966, and inConstruction Reports, Series C30-65S, Valueof New Construction Put in Place, 1962—65,January 1966.

Additional series have been derived byNBER by combining certain constructioncategories, shown separately in the source.These are indicated in the individual seriesdescriptions.

Seasonal adjustment is by the compilingagency, except for the following, which wereadjusted by NBER: (1) 1939—46 segment ofseries on new dwelling units, current dollars,(2) 1939—46 segment of all constant dollarseries, and (3) 1939—45 period for residentialadditions and alterations.

Current data for the series presented here(except those derived by NBER) are availablein Construction Reports, Construction Activ-ity, Bureau of the Census, C30 reports(monthly); and in Construction Review, Busi-ness and Defense Services Administration. Thecurrent dollar series except "Other PrivateNonfarm Nonresidential Building" are alsocarried in Survey of Current Business.

CONTENT: New construction as measuredby these series is defined by the Bureau of theCensus as the complete, original erection ofstructures and essential service facilities, aswell as major additions and alterations. Main-tenance and repairs, while considered part ofconstruction, are not included in these serieson new construction. The value of such work


is estimated separately. When added to thatof new construction, it yields estimates of thevalue of total construction. (For annual dataon maintenance and repairs, see ConstructionStatistics, 1915—1964, and various issues ofConstruction Review.) Structures includebuildings; dams and silos; highways; road-ways; water and signal towers; electric lightand power transmission and distribution lines;petroleum and gas pipe lines and distributionlines; telephone and telegraph lines; radio,television, and radar towers; water supplylines; sewers; and all similar work which isbuilt into or affixed to the land. Service facil-ities considered part of construction are thosetypes of general purpose equipment which,when installed, become integral parts of thestructure, e.g., plumbing, heating, centrallighting and air-conditioning equipment, andelevators. Not included are special purposefixtures, such as steam tables, conveyor lines,home refrigerators, church pews, school lock-ers, and printing presses. Major additions andalterations include such projects as new wingsor stories, conversion of space to other uses,or the initial installation of fixed equipmentin existing structures. Construction coversclearing and development of the site, includ-ing any necessary demolition performed aspart of the construction contract. It alsoincludes force-account work, and work doneby owners or their families on homes or farmbuildings. It excludes oil and gas well drilling,digging and shoring of mines, and work con-sidered an integral part of farming operations.

The concept of "value of construction putin place" is defined as the value of workinstalled or erected at the site during a givenperiod. For a particular category of construc-tion, the figures represent the total value ofsuch work for all projects under way duringthe period, regardless of when work on eachproject was started. Value put in place isestimated as the sum of (1) the cost of con-struction materials actually consumed andlabor actually performed during the period,and (2) proportionate shares of charges for

construction equipment used, overhead costsof the project owners, contractor's profit onconstruction operations, costs of architecturaland engineering services, and miscellaneouscosts chargeable to the project. It excludesspeculative profits, sales costs, and cost of land.The estimates thus include only that part ofthe total construction cost of a project whichis proportional to actual construction activityduring a given period.2 In this respect theymay be thought of as measures of "expendi-tures," as differentiated from the value ofcontracts awarded or the value of housingstarts.

Geographic coverage was modified slightlyin 1959, with the inclusion of Alaska andHawaii. The addition of the two new statesraised the national totals by approximatelyone-half of 1 per cent. (The estimates for newprivate nonfarm housing units are to beinterpreted as including Alaska and Hawaiibeginning 1946.)

Definitions of the various categories intowhich new construction has been classifiedare included in the descriptions of the individ-ual series.

DERIVATION OF ESTIMATES: Estimates of"value put in place" for the various categoriesof construction are derived from a variety ofbasic sources, and different estimating pro-cedures are used for different types of con-struction. Over the period covered by theseries, there have been some changes in pro-cedures, but the general approach hasremained the same. An attempt has been madehere to supplement information on currentprocedures with references to the most impor-tant modifications.

The three general procedures for derivingestimates of "value put in place," in order oftheir importance as measured by dollar vol-ume, are: (1) use of progress patterns applied

2 For a broader definition of expenditures, includingspeculative builders' margins, see Goldsmith, A Studyof Saving, Vol. I, Table R.3O, p. 622, and The NationalWealth of the United States in the Postwar Period,Princeton for NBER, 1962, Tables B-i, p. 225, andB-3, p. 227.


to starts data, which in turn are derived fromcontract award or building permit data,supplemented in the case of residential build-ing by sample surveys; (2) imputing monthlyvalues from annual expenditure data obtainedfrom public utility companies, trade associa-tions, or public agencies; and (3) tabulationof data from progress reports of federalagencies, and •a few state and local govern-ment agencies supervising construction ofpublic works.

Estimating procedures for the different typesof construction are discussed in the individualseries descriptions.

Estimates of the physical volume, or valuein constant dollars, of new construction putin place are made by deflating the currentdollar series for each type of construction bya cost index.3 (See individual series descrip-tions for specific deflators used.) Constantdollar series which are aggregates are obtainedby adding constant dollar figures for the vari-ous components. The price indexes used forthe deflation of construction expendituresare essentially wage rates and building mate-rial prices combined into an index under theassumption that there has been no change inproductivity in construction. For a discussionof these indexes see The Price Statistics of theFederal Government, NBER, 1961, AppendixB, and "Notes on the Measurement of Priceand Quality Changes" by Zvi Griliches, inModels of Income Determination, Studies inIncome and Wealth, Vol. 28, NBER, 1964, pp.386—388.

REVISIONS: Preliminary estimates of valueput in place are prepared each month for thevarious categories of construction, and theseare later revised as more complete informationbecomes available from primary sources. Forsome categories the revisions are only slight,because most of the data involved are avail-able when the preliminary estimate is made.

3 For a description of the indexes used, see Con-struction Reports, Construction Activity, C30—61Supplement, Bureau of the Census, October 1964, pp.81—83.

For other series, however, the preliminaryand revised estimates may be significantlydifferent.4 About the end of May of each year,revisions are made, when indicated, in thefigures for the previous two years.

Over the period 1960 to 1963, in additionto such routine revisions, substantial changeswere made in the level of estimates for fourtypes of construction: new private housingunits, private residential additions and altera-tions, farm construction, and public utilities.These revisions were carried back to 1959,creating a break in comparability with datafor earlier years, both for these four categoriesand for the aggregate series of which they arecomponents. In 1963, all constant dollar seriesfor 1959 and subsequent years were changedfrom a 1947—49 to a 1957—59 base.

In 1964 further revisions were introducedwhich extended all series, both in current andconstant dollars, back to 1946 on a basis com-parable to the current series. See individualseries descriptions for further discussion ofthe revisions and the extent of comparabilitywith years prior to 1946.

The revised constant dollar series, beginningwith data for 1946, are available from thesource only in the form of seasonally adjusteddata at annual rates, in 1957—59 dollars. Theearlier segment for each of these series, cover-ing 1939—46, is available from the source onlyin the form of unadjusted data at monthlyrates, in 1947—49 dollars. We have derived aseasonally adjusted series at annual rates, in1947—49 dollars, for 1939—46, by applying theimplicit seasonal factors from the correspond-ing current dollar series, and multiplying by12. Conversion factors, for use in changingthe constant dollar values from a 1947-49 toa 1957—59 base, have been developed by theBureau of the Census. These are available inConstruction Reports, Construction Activity,C30—61 Supplement, p. 5.

REFERENCES: For more detailed descriptionsand evaluations of the Department of

4 See, for example, descriptions of series on privatepublic utilities, and on public construction.


Commerce series, including the historicaldevelopment of the present program, see:

Construction Reports, Construction Activ-ity, C30—61 Supplement, U.S. Bureau of theCensus, October 1964.

1961 Proceedings of the Business and Eco-nomic Statistics Section, American StatisticalAssociation, papers and discussions includedin "Statistical Coverage of the ConstructionIndustry: A Critique," pp. 168—190.

Elmer Clark Bratt, Appraisal of StatisticalIn formation on Construction in the U.S. (andAppendices), Office of Statistical Standards,U.S. Bureau of the Budget, 1957—1958.

Construction Volume and Costs, 1915—1956,Statistical Supplement to Volume 3 of Con-struction Review, Business and Defense Serv-ices Administration and Bureau of LaborStatistics, 1958.

Lowell J. Chawner, Construction Activityin the United States, 1915—37, Domestic Com-merce Series No. 99, Bureau of Foreign andDomestic Commerce, U.S. Department ofCommerce, 1938.

Individual Series

C21. Total New Construction,Dollars, Annual, 1915—1963;1939—1963.

C22. Total New Construction, ConstantDollars, Annual, 1915—1963; Monthly,1939—1963.

These include all new construction. Theyare the sum of private construction (SeriesC23 and C24) and public construction (SeriesC49 and C50). For definition and derivationof estimates, see general description.

C23. Private Construction, Current Dollars,Annual, 1915—1963; Monthly, 1939—1963.

C24. Private Construction, Constant Dollars,Annual, 1915—1963; Monthly, 1939—1963.

The distinction between private and publicconstruction is based on ownership, rather

than source of funds or financing. Theseseries therefore represent all privately ownedconstruction, including projects of nonprofitorganizations subsidized by federal, state orlocal governments, e.g., hospitals built underthe National Hospital Program. The categoriesof construction included are those coveredby the component series for private nonfarmresidential buildings (Series C25 and C26);private nonf arm nonresidential buildings(Series C33 and C34); farm construction (SeriesC41 and C42); public utilities (Series C43 andC44); and an "all other private" category (notpresented here as a separate series), whichincludes water supply dams and reservoirs(not constructed by public utility companies),waterfront improvements, sewer and waterfacilities, airfield construction other thanbuildings, parks and playgrounds, and othermiscellaneous nonbuilding items.

The procedures for deriving the current andconstant dollar estimates for the variouscategories of private construction are describedbelow under the individual component series.The current dollar estimates for "all otherprivate" are based on F. W. Dodge Companycontract award data in thirty-seven easternstates, and building permit data. The estimat-ing procedure is the same as for private non-residential buildings (see Series C33). Theconstant dollar figures for "all other private"are derived by using as a deflator theunweighted average of two monthly costindexes—Associated General Contractors, andEngineering News-Record.

C25. Private Non farm Residential Buildings,Current Dollars, Annual, 1915—1963;Monthly, 1939—1963.

C26. Private Non farm Residential Buildings,Constant Dollars, Annual, 1915—1963;Monthly, 1939—1963.

This category of construction is the sum ofnew housing units (as defined in Series C27and C28); residential additions and alterations(Series CM and C32); and a "nonhousekeep-

CurrentMonthly,


ing" category. The last is not presented hereas a separate series. For definitions and deriva-tion of estimates, see Series C27, C28, C3l,and C32.

The nonhousekeeping classification includesfixed structures, such as transient hotels,motels, dormitories, clubhouses, nurses' homes,etc. Data are estimated from F. W. DodgeCompany contract award data for thirty-seveneastern states, and from building permit data.The estimating procedure is the same as forprivate nonresidential buildings (see SeriesC33).

The constant dollar series for private non-farm residential buildings wasdeflating the current dollar figures by themonthly residential construction cost index ofE. H. Boeckh and Associates—an unweightedaverage of indexes for frame and brick resi-dences.

C27. Private Non farm New Housing Units,Current Dollars, Annual, 1915—1963;Monthly, 1939—1963.

C28. Private Non farm New Housing Units,Constant Dollars, Monthly, 1946—1963.

This category covers new nonfarm housesand apartments, at all levels of value andquality, which are housekeeping dwellings. Itincludes those built for permanent or sea-sonal occupancy. Excluded are mobile homes,new units created by remodeling or conver-sion of existing structures, and dwelling unitsin buildings which are primarily nonresiden-tial.

Estimates of the value put in place arederived from monthly estimates of the valueof housing starts. The starts data are con-verted to value-put-in-place estimates byapplying activity or progress patterns, whichindicate the percentage of the value of con-struction projects started which will be com-pleted in successive months. The derivationof the starts data begins with estimates of thenumber of units started, as described in Series

B14. The value of housing starts is then cal-culated by applying average construction costfigures, estimated separately for starts in per-mit-issuing areas and starts in nonpermitareas. For permit-issuing areas, cost estimatesare based on permit values, adjusted for under-statement and for cost of architectural andengineering work. For nonpermit areas, thecost estimates were based, until 1962, oninformation from field surveys. Currentlythey are based on an assumed linear relation-ship between costs in nonpermit areas andpermit valuations in permit-issuing areas.

The data on value put in place, beginningin 1946, reflect major revisions in the housingstarts' series introduced in 1960 and 1964 (seeSeries Bl4 for description of revisions). Thedata for 1946—63 therefore represent a con-tinuous series. There is a break in the seriesin 1946, since revisions were not carried backto prior years. However, the new publicationon Construction Statistics, 1915—1964 treatsthe series as continuous.

The current dollar series, C27, has beenextended back to 1889, and the data for 1915—20 modified by David M. Blank (see SeriesC58). Official monthly constant dollar dataare available for private nonfarm new hous-ing units only for 1946 and subsequent yearsbut could be extended back to 1939 becausethe deflating index, which is the same as forSeries C26, is available monthly. Official an-nual estimates for 1915—1945 in constant dol-lars, not available when these tables wereprepared, have now been published in Con-struction Statistics, 1915—1964. For a constantdollar series back to 1889 see Series C59.

C29. Private Non farm Housekeeping Resi-dential Buildings, Current Do liars,Annual, 1915—1963; Monthly, 1939—1963.

C30. Private Non farm Housekeeping Resi-dential Buildings, Constant Dollars,Monthly, 1946—1963.

These series have been compiled by NBERas the sum of private nonfarm new housing


units (Series C27 and C28) and residentialadditions and alterations (Series CS! and C32).For definitions and derivation of estimates,see descriptions of these component series.

C31. Private Nonfarm Residential Additionsand Alterations, Current Dollars, An-nual, 1915—1963; Monthly, 1939—1963.

C32. Private Non farm Residential Additionsand Alterations, Constant Dollars,Monthly, 1946—1963.

Residential additions and alterations con-sist of such projects as addition of floor areaor living space by finishing basements or attics,enclosing porches, or construction of a newwing or story; conversion of space to providean additional dwelling unit; new retainingwalls, walks, driveways and patios; generalremodeling or modernization by installingan additional bathroom, enlarging a kitchen,etc.; additional landscaping; initial installa-tion of a furnace, hot-water heater, electricalwiring, plumbing, air conditioning, etc.Excluded are appliances, such as refrigerators,stoves, and window air conditioners; andnormal maintenance and repair work.

Separate data on residential additions andalterations are not shown in the source for1960 and following years. Imputed values forthis category are, however, included in thefigures for private nonfarm residential build-ings. Data for 1960—63 have been derived hereby subtracting data for new housing unitsfrom private residential buildings.

A quarterly survey of owners and rentersof residential properties, conducted by theBureau of the Census from 1960 through 1963,is the basis for estimating the over-all level ofexpenditures for this type of construction.(This survey also covers expenditures forresidential maintenance and repairs. For adetailed description of the survey, see Con-struction Reports, Residential A iterations andRepairs, No. C50—1, June 1961, and for ananalysis of the quality of the survey, seeResponse Errors in Collection of Expenditures

Data by Household Interviews: An Experi-mental Study, Technical Paper No. 11, U.S.Bureau of the Census, 1965.) The averageannual outlay derived from the survey for the1960—62 period is imputed for each of theyears 1960—62. For 1963, the imputation isbased on the average annual outlay for theyears 1960—63. Monthly figures for 1960—63were derived by distributing the annual totalsaccording to seasonal indexes interpolatedfrom quarterly seasonal movements. "Oldseries" estimates for 1945 to 1959 had been pre-pared by the Bureau of Labor Statistics basedon a procedure developed in 1957. (For details,see Marvin Wilkerson, "Revised Estimates ofResidential Additions and Alterations, 1945—56," Construction Review,June 1957.) In 1964,the "old series" estimates for 1951 through1959 were revised to levels approximatelycomparable with the current series by linkingthe percentage difference between the old andnew series estimates for 1960 to the previousbase year—1950. For 1950 and prior years,data were considered comparable to thecurrent series without revision. Estimates foryears previous to 1945 were made in 1950 bythe BLS, primarily on the basis of its FamilyExpenditures Surveys.

The constant dollar series for private non-farm residential additions and alterations wasderived by deflating the current dollar figuresby the monthly residential construction costindex of E. H. Boeckh and Associates—anunweighted average of indexes for frame andbrick residences. Monthly data in constantdollars are available only from 1946 on, andonly in seasonally adjusted form.

C33. Private Non farm Nonresidentialings, Current Dollars, Annual,1963; Monthly, 193 9—1963.

C34. Private Nonfarm Nonresidential Build-ings, Constant Dollars, Annual, 1915—1963; Monthly, 1939—1963.

Includes private industrial buildings (SeriesC35 and C36), commercial buildings (Series

Build-1915—


C37 and C38), and other private nonresiden-tial buildings (Series C39 and C40).

Buildings are classified according to theirspecific function, rather than the function ofany broader facility of which they are a part.However, in the case of most complex indus-trial projects, it is not possible to separatedata for the various components. The entireproject in these cases is classified by its majorfunction.

The current dollar estimates of value putin place for each type of construction aredeveloped in three steps:

1. Estimation of the value of constructioncontracts awarded. The basic contract dataused cover thirty-seven eastern states and theDistrict of Columbia and are compiled by F.W. Dodge Company (see general descriptionof Dodge series in Section A). The thirty-seven-state Dodge contract data are adjustedto approximate the national total by applyingannual ratios of the value of building per-mits in the area covered to that in the totalU.S. These ratios are computed separately foreach major type of construction. Furtheradjustments are made for undercoverage ofprojects, and for architectural and engineeringfees.

2. Use of the adjusted value of contractsawarded in a given month as an estimate ofthe value of starts in the following month.

3. Application to the starts estimates of aprogress pattern, developed for the particulartype of building, which indicates the per-centage of total value that will be put inplace each month, taking into account theaverage size of project and the season in whichwork is begun.

The constant dollar figure for each periodis the sum of the deflated values for thatperiod in the three component series. Eachcomponent has a different deflator, which isdescribed under the individual series.

For estimates of "all other private" con-struction and institutional construction backto 1837, and an allocation between corporateand unincorporated business since 1897, seeGoldsmith, A Study of Saving, Vol. I, Tables

R-17, p. 604; R-l8, p. 605; R-26, p. 617; R-27,p. 619; and R-28, p. 620.

C35. Private Industrial Buildings, CurrentDollars, Annual, 1915—1963; Monthly,1939—1963.

C36. Private Industrial Buildings, ConstantDollars, Annual, 1915—1963; Monthly,1939—1963.

Included in this category are production,assembly, and warehouse buildings in allmanufacturing industries; refrigeration, ice,and cold storage plants of warehouses; indus-trial grain elevators and storage silos; drycleaning plants and laundries. In practice,the data may include auxiliary facilities inmanufacturing establishments (e.g., restau-rants, parking lots, sewers) although they areexcluded in concept.

For derivation of current dollar estimates,see Series C33.

The deflating index used for the constantdollar data is the annual cost index of TurnerConstruction Company, interpolated monthlyby the "Commercial and Factory" cost indexof E. H. Boeckh and Associates.

For estimates of unincorporated businessexpenditures for new industrial structures backto 1897, see Goldsmith, A Study of Saving,Vol. I, Table R-13, p. 599.

C37. Commercial Buildings, Current Dollars,Annual, 1920—1963; Monthly, 1939—1963.

C38. Commercial Buildings, Constant Dollars,Annual, 1920—1963; Monthly, 1939—1963.

This category is composed of two sub-groups, for which separate data are availablein the source for 1920 and following years.(For the years 1915—19, data on commercialbuildings are not separate from "other privatenonfarm nonresidential buildings," as definedin Series C39 and C40.) The two subgroupsare as follows.

1. Office Buildings and Warehouses, whichincludes office buildings, except those builtby public utility companies for their ownuse; loft buildings; banks and building and


loan association buildings; warehouses andstorage buildings (other than cold storagebuildings, grain elevators, storage silos, andwarehouses built by industrial or publicutility companies for their own use). Becauseof a change in the F. W. Dodge Corporation'sclassification, beginning with data for 1956,warehouses for commercial establishments notengaged in selling storage space (e.g., depart-ment store warehouses) are included, whileprior to that date they were excluded fromthis category and included in "miscellaneousnonresidential buildings." Estimates for thesecategories are therefore not strictly compar-able between the two periods.

2. Stores, Restaurants, and Garages, whichincludes all buildings and structures intendedfor use primarily in the wholesale, retail, andservice trades. For example, complete shoppingcenters, department stores, drug stores,restaurants, public garages, and auto servicestations are in this category.


The constant dollar series is derived byadding deflated values for each of the twosubgroups defined above. Office buildings andwarehouses data are deflated by the annualcost index of George A. Fuller Company,interpolated monthly by the "Apartment,Hotel, and Office" cost index of E. H. Boeckhand Associates. Data on stores, restaurants,and garages are deflated by the monthly costindex of the American Appraisal Company.

For estimates back to 1897 of noncorporateexpenditures for new commercial structures,see Goldsmith, A Study of Saving, Vol. I,Tables R-lO, p. 595, and R-1l, p. 596.

C39. Other Private Non farm NonresidentialBuildings, Current Dollars, Annual,1920—1963; Monthly, 1939—1963.

C40. Other Private Non farm NonresidentialBuildings, Constant Dollars, Annual,1920—1963; Monthly, 1939—1963.

These series represent the sum of five build-ing categories: religious, educational, hospital

and institutional, social and recreational, andmiscellaneous nonresidential. Separate datafor each group are available in the sourcesfor 1920 and following years. For the years1915—19, data on this category are not sepa-rable from "commercial buildings," as definedin Series C37 and C38. In general, this cate-gory covers all private nonfarm nonresiden-tial buildings not included in "industrial" or"commercial," except buildings constructedby railroads, local transit companies, andpublic utility companies for their own use.For a detailed listing of types of buildingsincluded, see Construction Reports, Cons truc-tion Activity, C30—6l Supplement, U.S. Bureauof the Census, October 1964, p. 73. Becauseof a change in the classification of one typeof warehouse from "miscellaneous nonresi-dential" to "commercial," the data beginningin 1956 for these two categories, and for"other private nonresidential buildings" arenot strictly comparable to those for the periodbefore that date (see Series C37 and C38 fordetails).


The deflating index used to derive the con-stant dollar series is the monthly constructioncost index of the American Appraisal Com-pany.

C41. Private Farm Construction,Dollars, Annual, 1915—1963;1939—1963.

C42. Private Farm Construction, ConstantDollars, Annual, 1915—1963; Monthly,1939—1963.

Included in this category are: (1) new hous-ing units and additions and alterations toexisting units on places classified as farms,according to the 1960 Census definition; and(2) other buildings and structures used in farmproduction, such as barns, storage houses,smoke houses, fences, wells, etc. Excludedare operations which are an integral part offarming, such as plowing, terracing, anddigging of drainage ditches.

CurrentMonthly,


Current dollar estimates of value put inplace are prepared separately for farm resi-den tial and nonresidential construction.Separate data for the two components areavailable on an annual basis only. The annualestimates are prepared by the Farm IncomeBranch of the Economic Research Service,U.S. Department of Agriculture. The currentestimates are based primarily on benchmarkdata from a 1955 sample survey of farm con-struction expenditures. Estimates for otheryears are interpolated and extrapolated onthe basis of income and expenditure datafrom past benchmark years, and other relevanteconomic data available currently. (For fur-ther details, see Major Statistical Series, U.S.Department of Agriculture, "Gross and NetFarm Income," Volume III, December 1957.)Monthly estimates of farm construction, sea-sonally adjusted, are then prepared by theBureau of the Census by fitting a trend lineto the monthly averages of successive annualestimates. A constant seasonal pattern is thenapplied to derive the "unadjusted" monthlydata.

In 1964 the series was revised back to 1946,on the basis of revised annual estimates ofthe Department of Agriculture. There is abreak in comparability in 1946, pending exten-sion of the revisions to years prior to thatyear.

Earlier data, in million dollars (see below),are published in the U.S. Department of Agri-culture publication, Income Parity for Agri-culture, Part II, Section 5, 1941, p. 2.

Operators' ServiceDwellings Buildings

1910 185 2501911 175 2401912 190 2601913 190 2601914 190 2601915 195 270

Estimates of farmers' expenditures forresidential, and other buildings, back to

extrapolated on the basis of a fixed percentageof gross farm income, are available in Gold-smith, A Study of Saving, Vol. I, Table A-7,p. 761. A rough extrapolation back to 1837,on the basis of the number of farms and theincrease in the number of farms, is availablein the same source, Tables R-26, p. 617, andR-27, p. 619.

Estimates for the residential and nonresi-dential components are deflated separatelyand added together to obtain the constantdollar series. The annual farm constructioncost index of the Agricultural Marketing Serv-ice is used for both components. Monthlyfigures for farm housing units are interpolatedby the Boeckh residential index, and forother farm construction by the AmericanAppraisal Company index.

C43. Private Public Utility Construction,Current Dollars, Annual, 1915—1963;Monthly, 1939—1963 -

C44. Private Public Utility Construction,Constant Dollars, Annual, 1915—1963;Monthly, 1939—1963.

This category consists of six subgroups:railroads, telephone and telegraph, electriclight and power, gas, petroleum pipe lines,and, prior to 1959, local transit. (Beginning1959, local transit company expenditures forbuildings are included in the appropriatecategory under "nonresidential buildings,"and expenditures for nonbuilding projectsare in the "all other private" category.) Sep-arate annual data for these subgroups areavailable in the source. Separate monthly dataare shown only for telephone and telegraph(1939 and following) and for railroads (1939—45).

In general, the private public utility con-struction series include all private capitalexpenditures in public utility industries forbuildings or outside plant, and excludemachinery and equipment. They thereforeinclude not only types of construction peculiarto the operation of the type of utility, but also

Total435415450450450465

total,1897,


other types of nonresidential constructionbuilt by these companies for their own use.For a detailed list of types of projects covered,see Construction Volume and Costs, 1915—1956, Statistical Supplement to ConstructionReview, p. 72.

Current dollar estimates for each industryare based on expenditure data taken from thefiscal records of the utility companies. Thesedata are compiled by federal regulatoryagencies or cooperating private organizations.They are adjusted by the compilers for under-coverage and to eliminate nonconstructionitems.

Monthly construction expenditures dataare available only for the telephone and tele-graph component. The American Telephoneand Telegraph Company reports monthly con-struction expenditures by Bell System com-panies, with estimates for independent com-panies. The Western Union Telegraph Com-pany reports its monthly construction expendi-tures.

For railroads, electric light and power, gas,and petroleum pipelines, sources of data are:the Interstate Commerce Commission (forrailroads and petroleum pipelines); the Fed-eral Power Commission and the Rural Elec-trification Administration (for annual esti-mates on electric light and power); the EdisonElectric Institute (for preliminary quarterlyestimates on electric light and power); and theAmerican Gas Association.

Data for local transit expenditures prior to1955 were based on annual estimates compiledby the American Transit Association. For1955—58 they are projections based on anassumed gradual decline from the 1954 levelto zero in 1959.

Revisions in the public utilities constructionseries, resulting from changes in estimatingprocedures for all categories except telephoneand telegraph, were introduced by the Bureauof the Census in 1963 and carried back to1959. In 1964 these revisions were extendedback to 1946, resulting in a break in the seriesin that year. However, the new publication

Construction Statistics, 1915—1964 treats theseries as continuous. The 1946—49 revisionswere not based on actual data from the pri-mary sources, but on assumptions drawn fromthe revisions for later years.

The constant dollar series is the sum of theindividually deflated component series. Valuesfor railroads, local transit, and (prior to 1946)telephone and telegraph are deflated by anappropriate Interstate Commerce Commissionannual cost index, interpolated monthly bythe Engineering News-Record index. Begin-fling with data for 1946, the deflator forthe telephone and telegraph component isan annual cost index of the American Tele-phone and Telegraph Company. Currentmonthly deflators are extrapolated from theAT&T index by the Engineering News-Recordindex. The monthly data are revised by inter-polation from the succeeding year's AT&Tindex, when this becomes available. Electriclight and power expenditures are deflated bya weighted average of two Handy-Whitmansemi annual indexes—electric plant, weighted9, interpolated monthly by ENR index; andutility building, weighted 1, interpolatedmonthly by Boeckh's commercial and factoryindex. Gas expenditures are deflated in thesame manner, except that an index on gasplant instead of electric plant is used. Petro-leum pipe lines data are deflated by anunweighted average of: (1) Handy-Whitmanindex on electric plant, interpolated by ENRindex, (2) Handy-Whitman index on gas plant,interpolated by ENR index, (3) Handy-Whit-man index on utility building, interpolatedby Boeckh's commercial and factory index,and (4) ICC railroad annual cost index, inter-polated by ENR.

For rough estimates back to 1897 of publicutility construction expenditures, derived asa fixed percentage of total commercial, indus-trial, and public utility construction, see Gold-smith, A Study of Saving, Vol. I, Table R-16,p. 603.

Additional estimates of gross capitalexpenditures for "all regulated industries,"


with a breakdown into steam railroads, elec-tric light and power, telephones, street andelectric railways, local bus lines, and all other,have been developed by Melville Ulmer. Thesemeasures include plant and equipment. SeeMelville J. Ulmer, Capital in Transportation,Communications, and Public Utilities: ItsFormation and Financing, Princeton Univer-sity Press for NBER, 1960.

C45. Non farm Building Construction, CurrentDollars, Annual, 1915—1963; Monthly,1939—1963.

C46. Non farm Building Construction, Con-stant Dollars, Annual, 1915—1963;Monthly, 1939—1963.

These series have been compiled by NBERas the sum of private nonfarm residentialbuildings (Series C25 and C26), private non-farm nonresidential buildings (Series C33 andC34), public nonfarm residential buildings,and public nonresidential buildings. Data onthe latter two categories are available in theBureau of the Census source material but arenot presented here as separate series. Theaggregate series on nonf arm building construc-tion include all nonfarm buildings with thefollowing exceptions: (1) private public utilitybuildings; (2) all buildings on military installa-tions, excluding federal industrial installationsand housing under the armed services housingprograms; (3) buildings connected with stateor local public service enterprises or federalconservation and development projects. Theyalso include nonbuilding items associated withfederally owned industrial plants, such asthose constructed by the Defense Plants Cor-poration, the Department of Defense, and theAtomic Energy Commission. For a descriptionof the derivation of the estimates for thecurrent and constant dollar values, see seriesC25 and C26; C33 and C34; and C49 and C50.

C47. Commercial, Industrial, and Public Util-ity Construction, Current Dollars, An-nual, 1920—1963; Monthly, 1939—1963.

C48. Commercial, Industrial, and Public Util-ity Construction, Constant Dollars, An-nual, 1920—1963; Monthly, 1939—1963.

These series have been compiled by NBERas the sum of commercial buildings (SeriesC37 and C38), private industrial buildings(Series C35 and C36), and private public util-ity construction (Series C43 and C44). Fordefinitions and derivation of estimates, seedescriptions of these component series. Forrough estimates of these series back to 1837,see sources mentioned in notes to Series C33and C34.

C49. Public Construction, Current Dollars,Annual, 1915—1963; Monthly, 1939—1963.

C50. Public Construction, Constant Dollars,Annual, 1915—1963; Monthly, 1939—1963.

Included in these series are all new con-struction projects owned by federal, state, orlocal government. Privately owned projectsfinanced or subsidized by government fundsare excluded. These series are aggregates ofthe following categories, for which separatedata are presented in the source: residentialbuilding; nonresidential building (includingsubcategories of industrial, educational, hos-pital and institutional, administrative andservice, and other nonresidential); militaryfacilities (Series C55 and C56); highways(Series C51 and C52); sewer and water systems(shown separately and combined); publicservice enterprises; conservation and develop-ment; and "all other public." For a descrip-tion of the classification scheme and a list oftypes of projects included, see ConstructionReports, Construction Activity, C30—6 1Supplement, U.S. Bureau of the Census,October 1964, pp. 79—80.

Estimating procedures for the various typesof public construction are developed accordingto the kind of basic data available. For mostfederally owned projects and for a few cate-gories of state and local construction (e.g.,New York City Housing Authority), monthlyprogress reports are provided by the agencies


administering the construction programs.Normally, the progress reports are based onthe engineer's evaluation of the physical prog-ress of the project. In some instances theyrepresent expenditures rather than work putin place. Where this kind of information isnot available, the general method is to esti-mate construction starts from contract awarddata, adjusted for omissions and duplicationsand for architectural and engineering costs—then to convert starts to value-put-in-placeestimates by using progress patterns. The con-tract award data are obtained from the F. W.Dodge Corporation reports for thirty-seveneastern states, and from various constructionnews services in the western states. This pro.cedure is used for most state and locally ownedprojects, except highways, and for some fed-eral projects. In converting contract awarddata to work-put-in-place estimates, a progresspattern is normally applied to the sum ofmonthly contract award for a particular typeof construction. However, in some cases sufil-cient information is obtained on individualprojects, so that estimates of value put in placeare developed for each project on the basis ofan appropriate progress pattern for that proj-ect. The latter procedure is used for the newhousing units component of the state and localresidential category, for Capehart and Wherryarmed services housing, and for Navy indus-trial facilities. For a few relatively minor fed-eral construction programs, value-put-in-placeestimates are based on information in thefederal budget document, rather than onreports from individual agencies.

The constant dollar series is the sum of theseparately deflated components. Residentialbuilding values are deflated by the E. H.Boeckh residential index; industrial build-ings by the Turner Construction Companyannual index, interpolated monthly byBoeckh's commercial and factory index; andother nonresidential buildings by the Amer-ican Appraisal Company monthly index. Dataon military facilities are deflated by anunweighted average of (1) the American

Appraisal Company monthly index, (2) theBureau of Public Roads quarterly index, inter-polated monthly by the Engineering News-Record index, (3) the George A. Fuller Com-pany annual index, interpolated monthly byBoeckh's index for apartments, hotels andoffices, and (4) the Turner Construction Com-pany annual index, interpolated monthly byBoeckh's commercial and factory index. Dataon public service enterprises are deflated by aweighted average of two Handy-Whitmansemiannual indexes—electric plant (weighted9), interpolated monthly by the EngineeringNews-Record index; and utility building(weighted 1), interpolated monthly by Boeckh'scommercial and factory index. Data on sewerand water systems, conservation and develop-ment, and all other public construction, excepthighways, are deflated by an unweighted aver-age of two monthly indexes—Associated Gen-eral Contractors, and Engineering News-Rec-ord.

See Series C51 and C52 for estimating pro-cedures for highway construction. See alsoSeries C53 and C54 for annual data on pub-lic construction, by type of ownership. Forrough estimates back to 1837 of expendituresfor new public construction by U.S. and stateand local governments, see Goldsmith, A Studyof Saving, Vol. I, Tables R-26, p. 617, andR-27, p. 619.

C51. Highway Construction, Current Dollars,Annual, 1915—1963; Monthly, 1939—1963.

C52. Highway Construction, Constant Dollars,Annual, 1915—1963; Monthly, 1939—1963.

These series cover public construction ofstreets, roads, alleys, bridges, vehicular tunnels,viaducts, sidewalks, curbs, and gutters; newculverts and extensions of old ones; right-of-way drainage, erosion control, lighting, andguard rails; and earthwork protective struc-tures in connection with road improvements.

The annual value-put-in-place data repre-sent annual expenditure estimates preparedby the Bureau of Public Roads. Monthly data


on value put in place for federal-aid highways(interstate and state) are based on BPRmonthly estimates, adjusted to the level of theannual expenditure estimates for this cate-gory of highways. Monthly data for all otherhighways are based on the annual expenditureestimates, distributed monthly by applyingseasonal indexes computed from final data forfederally aided state highways. The final BPRannual estimates do not become availableuntil two years after the period reported. Datapublished before that time are preliminaryfigures based on projections made by the BPR.The procedure for estimating value of highwayconstruction put in place was revised in 1963.In 1964 data were revised back to 1946. Therevised data are considered comparable withdata for earlier periods.

The constant dollar series is derived from aquarterly cost index of the Bureau of PublicRoads, interpolated monthly by the Engineer-ing News-Record index.

See Series C20 for data on highway construc-tion by state and local governments prior to1915, and for references to similar Goldsmithseries extending back to 1897.

C53. Public Construction, Federally Owned,Annual, 1915—1963.

C54. Public Construction, State and LocallyOwned, Annual, 1915—1963.

SOURCE:19 15—1946: Construction Volume and Costs,1915—1956, Statistical Supplement to Volume 3of Construction Review, 1958. 1947—1963:Value of New Public Construction Put inPlace, by Sources of Funds, Ownership andType of Construction, 1 947—1963 Revised,November 1964.

These two series represent a breakdown byownership of the official series on total publicconstruction (Series C49). The source also pre-sents a breakdown by source of funds and across classification by ownership and type of

construction, annually (for 1915—1946) andmonthly, unadjusted for seasonal variation (for1947—1963). The current monthly series areavailable in Construction Review.

Federally owned construction (Series C53)covers projects on federal property financedwith federal funds. State and locally ownedconstruction (Series C54) includes projectsfinanced entirely by state and local govern-ments, as well as those financed in part byfederal grants-in-aid. Both series include workdone either by force-account or by contract.Beginning with data for 1959, Alaska andHawaii are included.

Estimating procedures for value of new pub-lic construction put in place are described inSeries C49. For a more detailed description,see source.

C55. Military Facilities,Annual, 1915—1962.

Current Dollars,

C56. Military Facilities, Constant Dollars,Annual, 1915—1962.

These series represent one component of theofficial series on new public construction(Series C49 and C50). Monthly data are avail-able in the source for 1939—62.

The data include all federally owned con-struction projects at military establishmentsexcept for: (1) Capehart housing, and rehabil-itation of Wherry housing for the armedservices, (2) industrial facilities, and (3) theCorps of Engineers—Civil works program.Military construction by states, such asarmories and rifle ranges, is excluded. Typesof projects covered include troop housing,administration and training buildings, ware-houses, mess halls, recreation centers, educa-tional facilities, hospitals, airfields, missilesites, fleet facilities, communication networks,etc.

Estimates of value of construction put inplace are based on monthly reports from theagencies supervising construction; or, for pro-grams consisting of a number of small proj-ects, on annual data from the federal budget.


The constant dollar series (C56) was derivedby applying a deflating index as described inSeries C50. The 1915—46 figures are shown in1947—49 dollars; and the 1946—62 figures in1957—59 dollars. The series is continuous inthe source except for the change in the base.The values in 1947—49 dollars can be convertedto approximate estimates in 1957—59 dollarsby multiplying by a conversion factor of 1.353.

For later data, see Construction Statistics,1915—1964, pp. 7 and 12.

C57. Federal Nonmilitary Construction, An.nual, 1915—1962.

SOURCE: This series was computed byNBER as the difference between federallyowned public new construction (Series C53)and military facilities (Series C55).

MISCELLANEOUS ESTIMATES OF RESIDENTIAL CONSTRUCTION, ANNUAL

C58. Expenditures for New Private Non farmHousekeeping Units, Blank, CurrentDollars, 1889—1921.

C59. Expenditures for New Private Non farmHousekeeping Units, Blank, 1929Dollars, 1889—1946.

SOURCE: David M. Blank, The Volumeof Residential Construction, 1889—1950, Tech-nical Paper 9, New York, NBER, 1954, Table18.

The concept of expenditures in these series,and the type of Construction covered, are thesame as those defined in the Commerce Depart-ment series on "value of new construction putin place, private nonfarm residential newhousing units." (See descriptions of SeriesC27 and C28.)

The two series as presented in the sourcecover the period 1889—1950. The currentdollar estimates are shown here only through1921, since for that year and subsequent yearsthe official estimates of the Bureau of LaborStatistics and the Department of Commercewere incorporated in the Blank series. Theseofficial estimates, including subsequent revi-sions for 1946 and following, are shown inthis volume as Series C27. The Blank constantdollar series (value in 1929 dollars) is shownthrough 1946, since, when these tables wereprepared, official estimates in constant dollarswere not available for prior years. (The 1946figure in 1929 dollars corresponds to the

current dollar figure for that year before the1964 Census revisions.) Series C28 shows therevised Bureau of the Census estimates in1957—59 dollars for 1946—63; estimates for1915—45 were published, too late for inclusionhere, in Construction Statistics, 1915—1964.Aside from the difference in the base periodsused, the Blank constant dollar series is notcomparable to Series C28 because of the 1946break in the current dollar estimates.

Blank's estimates were intended primarilyto extend the official current dollar series,which begins in 1915, to earlier years. Aftercomparing his 1915—29 estimates with theofficial series, Blank presented his data for1915—20 as an improvement over the officialseries, and accepted the official series for 1921and following years, since the differences inthese years were small. Although Blank'sestimating procedure was, in general, the sameas that used in the BLS-Census series, morecomprehensive source material and morerefined techniques were used for the Blankestimates. (See source, Section 4, for details.)

Blank estimated the permit value of newconstruction from BLS-WPA permit data bythe same methods as described for housingstarts (see Series Bl4, General Description),except that the calculation for all nonfarmdwelling units had to take account of the factthat rural nonfarm units had lower averagevalues than urban nonfarm units. The permitvaluations were then increased to allow forundervaluation of construction costs and to


cover architects' and engineers' fees, landdevelopment costs, and builders' profit mar-gins on construction operations. They werethen converted to a construction-put-in-placebasis, by extending the carryover of construc-tion from one year to the next.

The expenditure series does not include"speculative profits of operative builders" orclosing costs on new houses. Some estimatesof these amounts can be found in RaymondW. Goldsmith, The National Wealth of theUnited States, pp. 225—227, for 1945—58, andA Study of Saving, Vol. I, Table R-30, p. 622.

The constant dollar series was derived bydeflating the current dollar data by a con-struction cost index developed by Blank. Theindex for 1910—50 is based on the Boeckhindex of residential construction. Earlieryears are extrapolated by indexes of building-trade wage rates and building materials prices(for details, see source, Table 21; for referenceto Goldsmith estimates back to 1837, see thenote to Series CS and C4).

These Blank estimates were included in thestudy by Leo Grebler, David M. Blank, andLouis Winnick, Capital Formation in Residen-tial Real Estate: Trends and Prospects, Prince-ton University Press for NBER, 1956. For adiscussion of the merits of the data and con-clusions presented in this work and repliesby Grebler, Blank, and Winnick, see thefollowing articles: Margaret G. Reid, "CapitalFormation in Residential Real Estate," Jour-nal of Political Economy, April 1958, pp. 131—153; Grebler, Blank, and Winnick, "OnceMore: Capital Formation in Residential RealEstate," ibid., December 1959, pp. 612—619;Margaret Reid, "Reply," ibid., pp. 619—626;Grebler, Blank, and Winnick, "Rejoinder,"ibid., pp. 626—627.

C60. Estimated Production of Non farm Resi-dential Housekeeping Units, Gottlieb,1810—1939.

SOURCE: Manuel Gottlieb, Estimates ofResidential Building, United States, 1840—

1939, New York, NBER, Technical Paper 17,1964.

The data represent estimates of the numberof new nonfarm housekeeping units erectedeach year. They are presented as a suggestedrevision of the official estimates for the period1890—1939, and an extension of this seriesback to 1840. The estimates are intendedprimarily for use in the study of long-swingmovements of residential construction, and,according to the author, may not do justice toshort cyclical movements.

The procedure was first to establish decadetotals and then to interpolate annual move-ments for each decade. Decade estimates forthe period 1890—1940 are described below inconnection with the annual data. For theperiod 1860—90, the decade totals were esti-mated from Ohio data on number and valueof new buildings projected intonational aggregates. The estimates for the U.S.were derived from Ohio data in two ways.The first involved multiplying the U.S. decadeincrements in the stock of nonfarm occupieddwelling units (from the 1890 and 1910 Pop-ulation Censuses), by the ratio, for Ohio, ofthe number of new dwelling units built to theincrement in the stock of nonfarm dwellingunits. The second method involved multiply-ing the U.S. labor force increment in thedecade by the Ohio ratio of dwelling unitsbuilt to the labor force increment. The twovariants were then averaged and the resultinglevel of decade estimates was adjusted upwardby 1.58 per cent to fit the previously acceptedestimates of 1900—20. For the decade of the1850's, the estimate was based mainly onCensus data of stocks of dwelling units. Thedecade total for the 1840's was roughly esti-mated, primarily from data in the 1840 Census

5 For a description of the Ohio data, see ManuelGottlieb, "Building in Ohio Between 1837 and 1914,"Output, Employment, and Productivity in the UnitedStates After 1800, Studies in Income and Wealth 30,New York, NBER, 1966.


of Manufactures on number of dwellingserected in that year.

In order to derive annual figures for 1840—59, decade totals were distributed on the basisof decade indexes made up of the geometricaverage of Riggleman-Isard (Series B22) andan Ohio building value index prepared byGottlieb. Annual figures for 1860—89 werederived by distributing the decade totals bythe geometric average of three indexes equallyweighted: Gottlieb's Ohio index; Riggleman-Isard, deflated by a Riggleman cost-of-buildingindex; and Long's index of number of resi-dential buildings (Series B6). For 1890—99,Blank's yearly estimates (Series B14) werereduced by 15.3 per cent. Chawner's 6 decadeestimates were used for 1900—19 and weredistributed by index relatives made up of a.simple average of the Blank and Chawnerseries. Blank's "new series" (slightly higherthan the official series) was used for 1920—29,except that the 1920 and 1929 figures wereadjusted upwards for linking purposes. The1930—39 figures are official BLS estimates,8raised by 47 per cent to reach the level indi-cated by the "vintage" report of the 1940Census of Housing.

C61. Expenditures for Additions and Altera-tions, One- to Four-Family Non farmDwellings, Goldsmith, Current Dollars,1867—1914.

C62. Expenditures for Additions and Altera-tions, One- to Four-Family Non farmDwellings, Goldsmith, 1929 Dollars,1897—1914.

SOURCE: Raymond W. Goldsmith, A Studyof Saving in the United States, Vol. I, Prince-

6 Lowell J. Chawner, Residential Building, HousingMonograph Series, No. 1, prepared for NationalResources Committee, Washington, D.C., 1q39, p. 13.

David M. Blank, The Volume of Residential Con-struction, p. 59.

8 Construction During Five Decades, HistoricalStatistics, 1907—52, BLS Bulletin 1146, 1953, p. 3.

ton, 1955, Tables R-4, p. 587; R-5, p. 588; andR-31, p. 623.

The current dollar estimates were derivedby multiplying Goldsmith's construction esti-mates by the ratio of expenditures for altera-tions and additions to expenditures for newresidential construction. This ratio is extrap-olated back from 1920—30 (estimated at 10per cent on the basis of Department ofCommerce estimates) by an index of the move-ment of the ratio in Long's estimates. SeeClarence D. Long, Jr., Building Cycles andthe Theory of Investment. (A description ofthe Long series, B1—B9, is included in thisvolume.)

The constant dollar series was derived byapplying a building cost index to the currentdollar estimates. (See Table R-20, pp. 607—609, in the source.)

C63. Expenditures for Additions and Altera-tions, Housekeeping Units, Gre bier,Blank, and Winnick, Current Dollars,1889—1953.

C64. Expenditures for Additions and Altera-tions, Housekeeping Units, Gre bier,Blank, and Winnick, 1929 Dollars, 1889—1953.

souRcE: Leo Grebler, David M. Blank,and Louis Winnick, Capital Formation inResidential Real Estate: Trends and Prospects,Princeton University Press for NBER, 1956,pp. 40—52, 336, and 362.

Additions and alterations are defined as inthe BLS-Commerce series (Series C3l andC32). Official BLS-Commerce estimates werecarried back from 1921 by a crude graphicextrapolation using the movement of house-keeping construction expenditures but reduc-ing the amplitude of fluctuation.


INVESTMENT IN STRUCTURES, INCLUDING OIL AND GAS WELLDRILLING, ANNUAL AND QUARTERLY

C65. Total New Construction Expenditures,and Oil and Gas Well Drilling, BLS-Commerce, Annual, 1939—1963.

SOURCE: Sum of Series C21, value of totalnew construction put in place; and Series C84,expenditures for petroleum and natural gaswell drilling. The resulting series matches incoverage Kuznets' estimates of "Gross TotalConstruction," 1889—1955 (Series Cl), butincludes more recent data from the officialsources. For derivation of the estimates, seecomponent series.

C66. Gross Private Domestic Investment inStructures, Annual, 1929—1963, OBE;Quarterly, 1921—1939, Barger and Klein;Quarterly, 1939—1963, OBE.

C67. Gross Private Domestic Investment inStructures Other than Residential Non-farm, OBE, Annual, 1929—1963; Quar-terly, 1939—1963.

SOURCE:Annual, 1929—1963: Survey of Current Busi-ness, U.S. Department of Commerce, Office ofBusiness Economics, August 1965, Table 1.Quarterly, 1921—1938: Unpublished estimates,Barger and Klein. Quarterly, 1939, Barger andKlein segment: National Income Supplementto Survey of Current Business, July 1947,Table 43. Quarterly, 1939—1945, OBE oldseries: National Income, 1954 Edition, ASupplement to the Survey of Current Business,Table 45. Quarterly, 1946, OBE old series:U.S. Income and Output, a Supplement to theSurvey of Current Business, 1958, Table 1-3.Quarterly, 1946—1963, OBE revised series:Survey of Current Business, August 1965,Table 1.

Income and Product Accounts, prepared bythe National Income Division, Office of Busi-ness Economics, U.S. Department of Com-merce. Major revisions of these accounts in1965 provide estimates of fixed investment innonresidential, nonfarm residential, and farmresidential structures, annually from 1929, andquarterly from 1946 to the present. Series C66includes all three types of structures; SeriesC67 is the sum of nonresidential and farmresidential. The quarterly data are availableonly on a seasonally adjusted, annual ratebasis. In the revised estimates, fixed invest-ment in structures includes (1) the value ofnew private construction put in place, asestimated by the Bureau of the Census (seeSeries C23); (2) crude-petroleum and natural-gas well drilling (see Series C84); (3) commis-sions on the sale of structures; and (4) nettransfers of used structures from or to govern-ment.

Prior to the 1965 revisions, OBE estimateswere prepared for a "new construction" com-ponent of gross private domestic investmentwhich did not include the sales commissionsand net transfers components and which werebased on earlier estimates of the value of newconstruction put in place. These data, whichextended back to 1939 on a quarterly basis,are included here in Series C66 and C67 forpurposes of historical comparison. A one-yearoverlap in 1946 is provided. The break incomparability in the series at this point repre-sents differences in estimating procedures aswell as definitional changes, as explainedabove.

The quarterly series for total new construc-tion (C66) can be further extended to 1921on a seasonally adjusted basis by using esti-mates prepared by Harold Barger and Law-rence R. Klein for their article "A QuarterlyModel for the United States Economy," Jour-nal of the American Statistical Association,

These series represent thestructures component of grosstic investment included in

investment-in-private domes-the National


September 1954, pp. 413—437. These data weredeveloped from a series presented by HaroldBarger in Outlay and Income in the UnitedStates, 1921—1938, New York, NBER, 1942.The quarterly series on "private construction,"seasonally adjusted, in that publication wasrevised by Barger and Klein to conform to thelevel of the OBE annual series for the period1929—1938 published in National IncomeSupplement to Survey of Current Business,July 1947. Annual ratios of the OBE estimatesto the Barger estimates, 1929 to 1938, wereinterpolated to derive quarterly raising factorsfor this period. The raising factor for the firstquarter of 1929 was used for the entire period1921 to 1928. The resulting quarterly seriesfor 192 1—1938, presented here, was compara-ble to the OBE quarterly series for 1939—1946,

published in 1947. The original Barger quar-terly series underlying the Barger-Klein revi-sion was derived from annual Department ofCommerce estimates. Those for 1921—81, byLowell J. Chawner, appear in ConstructionActivity in the United States 1915—37, U.S.Dept. of Commerce, 1938. The estimates for1932—37, by Samuel J. Dennis, were publishedin Survey of Current Business, August 1939;those for 1938, in ibid., February 1940. Thequarterly movements were obtained by inter-polation from Dodge contracts data, laggedtwo months.

The cyclical analysis shown for Series C66is limited to the period beginning 1945. Thefigures used are from the 1965 OBE revisionsfor 1946 and later years, extrapolated back to1945 by the old OBE series.

ABRAMOVITZ INDEXES OF URBAN BUILDING ANDPUBLIC UTILITIES CONSTRUCTION

General Description

SOURCE: Moses Abramovitz, Evidences of LongSwings in Aggregate Construction Since theCivil War, New York, NBER, 1964.

Abramovi tz constructed three aggregateindexes for the period from 1870 to 1918 onthe basis of existing series measuring construc-tion activity, or some closely related variable,such as permits or contracts. These indexescover a combination of urban residential,urban nonresidential excluding government,and railroad and public utilities construction.These categories accounted for 75 to 80 percent of total construction in the period cov-ered. Public building was excluded becausesome unknown fraction of public buildingwas exempt from permits and the series con-structed depend on permit data. However,two indexes use an adjusted form of theRiggleman index, which includes some permitsfor public building.

The three series may be thought of as a

single index of urban building and publicutilities construction, expressed in three var-iants: (1) an index of value in current prices,(2) an index of value in constant prices, and(3) an index of volume of physical units. Eachindex is divided into two overlapping seg-ments in order to take advantage of morereliable and comprehensive data in the periodafter 1889. Public utilities other than rail-roads are not represented in the first segmentof any of the indexes because the data areunsatisfactory. However, railroad expendi-tures in this period were about three timesas great as those of other public utilities. Rail-road and public utilities construction is repre-sented in the second segment of the first twoindexes by a series which includes expendi-tures for equipment. The third index includesonly railroads, since there were no satisfactorydata on physical units for public utilities.

In order to combine the components ofeach series (urban residential, nonresidential,and railroads and public utilities) weightsrepresenting their relative importance were


applied. In estimating the proper weight tobe assigned to each component, two sourceswere used: (1) Kuznets' estimates of total con-struction (Series Cl), and (2) Kuznets' dataon real estate improvements (based on wealthfigures), adjusted for depreciation by usingGoldsmith's data on average life of structuresin different industries. The source discussesthese indexes in more detail.

Individual Series

C68. Index of Urban Building and PublicUtilities Construction, Value in CurrentDollars, Annual (Segment I, 1870—97;Segment II, 1889—1918).

Segment I: Weighted average of the follow-ing two series, each expressed as a relative ofits average standing in 1870—1897 beforebeing combined.

1. Riggleman's index of value of buildingpermits, adjusted by Isard. Includesurban residential and urban nonresiden-tial building; see Series B22 of this vol-ume. Weight: 0.75.

2. Gross capital expenditures of railroadsin current dollars, Ulmer. Includes roadand equipment; see Melville J. Ulmer,Capital in Transportation, Communica-tions and Public Utilities: Its Formationand Financing, Princeton for NBER,1960, Table C-i. Weight: 0.25.

Segment II: Weighted average of the follow-ing three series, each expressed as a relativeof its average standing in 1889—1918 beforebeing combined.

1. Expenditures for new private nonfarmhousekeeping dwelling units, in currentdollars, Blank; see Series C58 of thisvolume. Weight: 0.41.

2. Long's index of value of nonresidentialbuilding; see Series B4 of this volume.Weight: 0.23.

3. Gross capital expenditures of all reg-ulated industries in current dollars,

Ulmer. Includes construction and equip-ment; see Ulmer, Capital in Transporta-tion, Table B-i. Weight: 0.36.

C69. Index of Urban Building and PublicUtilities Construction, Value in Con-stant Dollars, Annual (Segment I, 1870—97; Segment II, 1889—1918).

Segment I: Weighted average of the follow-ing two series, each expressed as a relative ofits average standing in 1870—1897 beforebeing combined.

1. Riggleman's index of value of buildingpermits, adjusted by Isard (Series B22of this volume), deflated by priceindex implicit in Kuznets' estimates oftotal new construction in current dollarsand in 1929 dollars (Series Cl and C2of this volume). Weight: 0.75.

2. Gross capital expenditures of railroadsin 1929 prices, Ulmer. See Ulmer, Capitalin Transportation, Table C-i. Weight:0.25.

Segment H: Weighted average of the follow-ing three series, each expressed as a relativeof its average standing in 1889—1918 beforebeing combined.

1. Expenditures on new private nonfarmhousekeeping dwelling units, in 1929prices, Blank. See Series C59 of this vol-ume. Weight: 0.41.

2. Long's index of value of nonresidentialbuilding (Series B4 of this volume),deflated by price index implicit in Kuz-nets' current dollar and constant dollarestimates of total new construction(Series Cl and C2 of this volume), lessprivate nonfarm residential (series C3and C4) and military construction.Weight: 0.23.

3. Gross capital expenditures of all reg-ulated industries, in 1929 prices, Ulmer.Includes construction and equipment;see Ulmer, Capital in Transportation,Table B-i. Weight: 0.36.


C70. Index of the Physical Volume of UrbanBuilding and Public Utilities Construc-tion, Annual (Segment I, 1856—97; Seg-ment 11, 1889—1918).

Segment I: Weighted average of the follow-ing three series, each expressed as a relativeof its average standing in 1856—1897 beforebeing combined.

1. Long's index of number of residentialbuildings; see Series B6 of this volume.Weight: 0.51.

2. Long's index of number of nonresiden-tial buildings; see Series B7 of this vol-ume. Weight: 0.23.

3. Rail consumption, American Iron and

Steel Institute; see Series D16 of this vol-ume. Weight: 0.26.

Segment II: Weighted average of the follow-ing three series, each expressed as a relativeof its average standing in 1889—19 18 beforebeing combined.

1. Number of new private nonfarm dwellingunits started, Blank; see Series B14 ofthis volume. Weight: 0.49.

2. Long's index of number of nonresidentialbuildings; see Series B7 of this volume.Weight: 0.27.

3. Rail consumption, American Iron andSteel Institute; see Series Dl6 of thisvolume. Weight: 0.24.

FEDERAL CONSTRUCTION EXPENDITURES, FWA-TREAS UR Y, A NNUAL

C71. Federal Construction Expenditures,Total Public Works, Fiscal Years 1791—1939.

SOURCE: Historical Statistics of the U.S.1789—1915, U.S. Bureau of the Census, SeriesH27, p. 169, and Series H33, p. 170.

Statistics were compiled by the FederalWorks Agency for 1920—39 and by the Depart-ment of the Treasury at the request of theFederal Works Agency for 1791—1919.

Construction was defined in the source(notes on Series H33—35, p. 162) as "the erec-tion of any new structural or nonstructuralimprovement to land, ships, and floatingequipment, additions to and complete replace-ment of existing works."

Repair comprises "work necessary for therestoration and preservation of structures,nonstructural improvements to maintain landand floating equipment in a sound and serv-iceable condition, and minor alterations."

Despite the definitions, the series doesappear to include certain expenditures thatwere not construction or repair. For example,the purchase of the Panama Canal for $50

million in fiscal year 1904 was included in"other."

The definitions of construction and repairare said by the source to be identical in thetwo segments of the series. The second seg-ment is described as covering only permanentfederal construction agencies, but this term isnot defined and it is not clear whether anydiscontinuity is introduced by this limitation.The two segments are treated as Continuousby Raymond Goldsmith (A Study of Saving,Vol. I, pp. 974—975), who appears to have hadaccess to the Public Works Administrationworksheets giving a more detailed breakdownthan the published data.

The components of this series are "militaryand naval" (Series C73), "public building"(Series C72), "rivers and harbors" (Series C74),and two other series not covered here, "recla-mation" and "other."

The BLS-Commerce series for federal pub-lic construction (Series C53) is not Continuouswith this one because the BLS-Commerceseries includes only new construction, notrepairs.

The source does not indicate the type of


year, but it is probably the federal. govern-ment fiscal year. From 1791 through 1842these ended December 31; from 1844 to date,on June 30. The 1843 year covered only sixthonths, January 1 to June 30. The fact thatthe construction total for 1843 was roughlyhalf that of the preceding and following yearssupports this assumption. For fiscal year datessee Annual Report of the Secretary of theTreasury on the State of the Finances, FiscalYear 1964, p. 402.

C72. Federal Construction Expenditures, Pub-lic Buildings, Fiscal Years 1 791—1919.

SOURCE: Historical Statistics of the UnitedStates, 1789—1945, Series H30, p. 169.

The BLS-Commerce series on federal non-residential building in Construction Volumeand Costs 1915—56, Table 4, p. 11, appears tobe continuous with this series, even thoughthe BLS-Commerce data purport to includeonly new construction put in place, not repairs.The BLS-Commerce calendar-year figures arevery close to averages of adjoining fiscal-yearfigures of this series. See also notes to SeriesC71.

C73. Federal Construction Expenditures,Military and Naval Construction, FiscalYears 1791—1919.

souRcE: Historical Statistics of the UnitedStates, 1789—1945, Series H28, p. 169.

Military and naval construction expendi-tures exclude naval vessels. The BLS-Com-merce data (Series C55) include only new con-struction put in place, not repairs, and are forcalendar years. See also notes to Series C71.

C74. Federal Construction Expenditures, Riv-ers and Harbors, Fiscal Years 1822—1919.

SOURCE: U.S. Bureau of the Census, His-torical Statistics of the United States, Colonial

Times to 1957, Series Q245. Statistics werecompiled by the Department of the Treasuryat the request of the Federal Works Agency.The following were the original sources (seeHistorical Statistics, p. 443):1822—1882: Statement of Appropriations andExpenditures for Public Buildings, Rivers andHarbors, Forts, Arsenals, Armories, and OtherPublic Works from March 4, 1789 to June 30,1882, U.S. Senate Ex. Doc. Vol. 7, No. 196,47th Congress, 1st Session (Treasury Depart-ment Doc. No. 373), pp. 521—522. 1883—1919:Federal Works Agency records (compiled fromTreasury Department accounts).

This series is also found in HistoricalStatistics, 1789—1945, Series H29, and is oneof the components of "U.S. Federal Construc-tion Expenditures, Public Works, Total,"(Series C71).

The data cover construction and repairwork, and include expenditures for flood con-trol as well as for rivers and harbors.

Data for 1920—57, compiled by the Corps ofEngineers of the U.S. Army, are available inthe source. For data covering navigation only,1936—62, see Historical Statistics, Continuationto 1962 and Revisions, Series Q245.

Data on rivers, harbors, and flood controlconstruction are included in the BLS-Com-merce series on "conservation and develop-ment." See Construction Volume and Costs1915—1956, Statistical Supplement to Vol. 3of Construction Review, 1958; and Value ofNew Construction Put in Place 1946—1963Revised, Construction Reports, ConstructionActivity, C30—61 Supplement, Bureau of theCensus, October 1964.

C75. U.S. Federal Construction Expenditures,Nonmilitary, 1791—1919.

SOURCE: This series was derived by NBERas the difference between total federal expendi-tures for public works (Series C7l) and militaryand naval construction (Series C73).


TRANSPORTATION AND PUBLIC UTILITIES, ANNUAL

C76. Miles of Railroad Built, Census-RailwayAge, 1830—1952.

SOURCE:1830—1879: Historical Statistics of the UnitedStates, Colonial Times to 1957, Bureau of theCensus, Series Q43. 1880—1952: Railway Age,Simmons-Boardman Publishing Corp., NewYork, January 1, 1938, and later annualreview issues.

These data refer to miles of first track com-pleted in a calendar year. It is not clear atwhat stage the building of a mile of road wasreported. If construction of a line took severalyears, a company may have reported eachyear, or might have waited until the wholeline was completed. Sometimes the year ofcompletion differed from the year in whichtraffic was first carried; in such cases the mile-age may have been assigned to either year.

The data for 1830—79 are originally from theTenth Census Report, Vol. IV, Report on theAgencies of Transportation in the UnitedStates. These data were compiled by theBureau of the Census from reports of railroadsknown to exist in 1880. They are limited tomileage still in operation in the census year1880.

Railway Age obtained its figures at annualintervals from individual railroads and fromstate railroad commissions.

For data on gross capital expenditure,including road and equipment, beginning in1870, see Melville J. Ulmer, Capital in Trans-portation, Communications, and Public Util-ities: Its Formation and Financing, PrincetonUniversity Press for NBER, 1960, p. 255 ff.

Railroad investment data for the pre-CivilWar period, beginning in 1828, are availablein Albert Fishlow, American Railroads andthe Transformation of the Ante-Bellum Econ-omy) Harvard University Press, 1965.

C77. Increases in Railroad Track MileageOperated, ICC, 1877—1963.

SOURCE:1877—1890: Historical Statistics of the UnitedStates, Colonial Times to 1957, Bureau of theCensus, Series Q17. 1891—1956: Ibid., SeriesQ48. 1957—1962: Ibid., Continuation to 1962and Revisions, Bureau of the Census, February1965, Series Q48. 1963: Transport Statisticsin the United States, Part 1, Interstate Com-merce Commission, Bureau of TransportEconomics and Statistics.

These data are year-to-year changes in totaloperated track mileage as published in theabove sources. Data on increases for the years1877—90 and 19 17—63, are for calendar years;those for 1891—1916 are for years ending June30. Data on operated track mileage are com-piled by the Interstate Commerce Commission.Data for 1876—90 are from Railway Statisticsbefore 1890, ICC Statement No. 32151 (mimeo-graphed), 1932. For these historical data theICC used the annual Poor's Manual of Rail-roads. Data for 1891—1954 are from ICC,Statistics of Railways in the United States.Later data are from Transport Statistics inthe United States, Part 1.

Operated track includes main tracks, yardtracks, and sidings. The data cover Class I,II, and III railroads. Prior to 1908 they alsoincluded switching and terminal companies."Circular" and unofficial roads are excluded.Alaska and Hawaii are included beginningwith data for 1960. Each company reportedits total operated track mileage; since twocompanies might operate the same track,there is some duplication. Data for differentyears are often not entirely comparablebecause of changes in accounting and report-ing methods.


C78. Capital Expenditures for Additions andBetterments, Roadway and Structures,Class I Railroads, ICC, 1921—1963.

SOURCE:

1921—1955: Railroad Transportation, A Statis-tical Record, Association of American Rail-roads, Bureau of Railway Economics, Decem-ber 1956. 1956—1963: Selected Income andBalance-Sheet Items of Class I Railroads inthe U.S., Bureau of Transport Economicsand Statistics, Interstate Commerce Commis-sion, Statement No. M-125, December issuesthrough 1961; No. Q-125, 4th quarter issuesfor later figures.

These data are collected by the Bureau ofTransport Economics and Statistics fromreports of individual Class I railroads. Switch-ing and terminal companies are included.Additions refer to new facilities, such as tracks,building and other structures; and additionsto these facilities. Betterments compriseimprovements to existing facilities such asthe application of heavier rail in tracks,strengthening of bridges by substitution ofheavier members, and construction of superiorfloors or roofs in buildings.

These data are also published in HistoricalStatistics of the United States, Colonial Timesto 1957, Bureau of the Census, Series Q105.

C79. Increase in Wire Mileage of WesternUnion Telegraph Company, Annual,1867—1962.

SOURCE: Statistical Abstract of the UnitedStates, Bureau of Census, issues of 1913, 1920,1935, 1943, 1947, 1952, 1957, 1963.

Data on wire mileage, published in theabove source, are based on annual reports of

the Western Union Telegraph Company.Increases were computed by subtracting theprevious year's mileage from the mileage inthe year reported. Increases for 1867—1913are for years ending June 30; data for 1914—1962 are for calendar years. Alaska and Hawaiiare excluded. The large increase in mileagein 1943 reflects the acquisition of Postal Tele-graph Companies.

C80. Increase in Telephone Wire Mileage,1881—1963.

SOURCE:1881—1936: American Telephone and Tele-graph Company. 1937—1963: StatisticalAbstract of the United States, Bureau of theCensus, issues of 1943, 1957, and 1962 to 1965.

Increases were derived from annual data onwire mileage in the above sources. The figurefor the previous year was subtracted from thefigure for the year being computed.

For 1881—1956, the annual figures refer towire mileage in all systems. They were esti-mated by the American Telephone and Tele-graph Company, except for four benchmarkyears (1922, 1927, 1932, and 1937) in whichBureau of the Census data were used. For thesegment 1956—1963, the annual figures werederived by adding wire mileage reported byBell Telephone Companies and by independ-ent companies. Data for the independentcompanies are based on operating reportssubmitted by each company to the U.S. Inde-pendent Telephone Association. However, notall such companies report. The data from BellTelephone Companies exclude drop and blockwire.

Because of technological improvements,increases in wire mileage do not completelyindicate the increase in telephone capacity.


SHIPBUILDING, ANNUAL

C81. Merchant Vessels Built and Documented,Gross Tonnage, Annual, 1797—1962.

SOURCE:1797—1957: Historical Statistics of the UnitedStates, Colonial Times to 1957, Bureau of theCensus, Series Q179. 1958—1962: HistoricalStatistics of the United States, Continuationto 1962 and Revisions, Bureau of the Census.

The basic sources used for compilation ofthis series were various annual issues of Mer-chant Marine Statistics (originally preparedby the Commerce Department and now issuedby the Bureau of Customs), supplemented byrecords of the Bureau of Customs.

This series covers documented merchantvessels of five gross tons or more that weregranted registers, enrollments and licenses, orlicenses as "Vessels of the United States."Documentation, which implies ownership byUnited States citizens, confers certain benefitsand privileges, such as that of engaging incoasting trade.

These data are reported in "gross tonnage,"which refers to space capacity rather thanweight. Currently one ton is considered equiv-alent to 100 cubic feet; before 1865, one tonequaled 95 cubic feet. For 1797—1834 data,years ended December 31; 1835—1842, Septem-ber 30; 1843—1940, June 30; and 1941—1962,January 1. Figures for 1835 and 1843 coveronly nine months. Figures for 1938—62 arenot comparable with those for earlier yearsas they represent only those vessels still inexistence as part of the merchant fleet at theend of the twelve-month period in whichthey were built. Therefore, they exclude thosevessels which were lost, sold to the U.S. gov-ernment, sold to aliens, or removed from themerchant fleet for any other reason, beforethe end of the period.

For a history and more detailed description,and a discussion of the limitations of the data,

see Historical Statistics of the United States,Colonial Times to 1957, pp. 438—441.

C82. Ships and Boats Built for Domestic Con-sum ption, Value in Current Dollars,Shaw, Annual, 1889—1937.

C83. Ships and Boats Built for Domestic Con-sum ption, Value in 1913 Dollars, Shaw,Annual, 1889—1937.

SOURCE: William Howard Shaw, Value ofCommodity Output Since 1869, NationalBureau of Economic Research, 1947, pp. 57,69, and 75—76. In addition to the annual databeginning 1889, estimates are given in thesource for 1869 ($11.5 million), and 1879($19.4 million).

The data represent value of work on vesselsof five gross tons and over. For the period1889—1919, estimates of value of output wereused as estimates of value destined for domes-tic consumption. (No adjustment was madefor exports and imports.)

The primary source for the 1869—1919 esti-mates was the Census of Manufactures,decennially 1869—1899 and quinquennially1899—1919. State statistical reports from Penn-sylvania, Massachusetts, and New Jersey wereused to estimate the intercensal years from1889 to 1919. For the period 1919—33 the dataare based on Kuznets' estimates. (CommodityFlow and Capital Formation, I, New York,NBER, 1938) adjusted to improve com-parability with earlier years. (Shaw's figuresinclude all work done; Kuznets' total includedwork done on completed vessels only.) After1933 the estimates are rough extrapolations.

Values in 1913 prices were derived by apply-ing price indexes to the current price series.For 1915—1937 the deflator was an InterstateCommerce Commission index for floating


equipment, published in a mimeographedrelease by the 1CC, Engineering Section,Bureau of Valuation, July 1, 1940. For 1889—1914 two indexes of cost to railroads of float-ing equipment were averaged. These indexes

were reported in Cost of Floating Equipment(President's Conference Committee, EasternGroup Pamphlet 290). The average was linkedto the index for later years.

For a detailed description, see source.

OIL AND GAS WELL DRILLiNG, ANNUAL

C84. Expenditures for Petroleum and NaturalGas Well Drilling, Dollars,Kuznets, 1889—1928; Commerce Depart-ment, 1929—1963.

SOURCE:

1889—1928: Unpublished annual estimateswhich underly five-year moving averages pub-lished, with explanatory notes, in Simon Kuz-nets, Capital in the American Economy,Princeton University Press for NBER, 1961,Tables R-5, pp. 492—493; R-15, pp. 526—527;and R-30, pp. 576—587. 1929—1938: NationalIncome, 1954 Edition, Supplement to Surveyof Current Business, Footnote 8 to Table 31,p. 209. 1939—1945: Construction and BuildingMaterials, Statistical Supplement, May 1954,U.S. Department of Commerce, Business andDefense Services Administration, Table 24,p. 56. 1946—1956: U.S. Income and Output,Supplement to Survey of Current Business,1958. 1957—1963: Survey of Current Business,July issues, 1962 and 1964.

The Kuznets data were calculated bymultiplying the series in 1929 prices (seeSeries C85) by the price index for petroleumpipe lines. For 1915—28, this price index wascalculated from Construction and BuildingMaterials, Statistical Supplement, May 1954,pp. 33 and 82, and it was extrapolated backfrom 1915 by the total construction cost indexdescribed in Kuznets, National Product Since1869, Table IV-4, notes to line 1.

The Commerce Department estimates wereprepared by the Building Materials and Con-struction Division of the Business and DefenseServices Administration. They represent allcosts of drilling, including the cost of casings.The cost of installed production equipment,

such as flowing and pumping equipment, isexcluded.

The. figures are based on the cost of drillingoil and gas wells, as reported in the Censusof Mineral Industries, 1939 and 1958, inter-polated and extrapolated by annual data onthe number of wells completed (from tradesources) and on average cost per well (esti-mated by the compiling agency).

For an extrapolation back to 1897 ofexpenditures for oil and gas well drilling, onthe basis of the value of output, see Gold-smith, A Study of Saving, Vol. I, Table R-33,p. 626.

C85. Expenditures for Petroleum and NaturalGas Well Drilling, Constant Dollars,Kuznets, '1889—1945; Commerce Depart-ment, 1946—1963.

SOURCE:1889—1945: Unpublished annual estimateswhich undérly five-year moving averages pub-lished, with explanatory notes, in Simon Kuz-nets, Capital in the American Economy.1946—1963: Same as above for current dollarseries (C84).

The Kuznets data, in 1929 dollars, for 1889—1918 were extrapolated from the 1919 figureby the number of wells drilled each year (seepp. 526—527 of Kuznets, Capital in the Ameri-can Economy for derivation). For 1919—28,data were extrapolated from the 1929 figureby the series described in Kuznets, NationalIncome and Its Composition, 1919—1938, p.645. This series is based on individual statedata for number of wells drilled and 1935costs. For 1929—45, Kuznets used the Corn-


merce series in current prices (C84) dividedby the price index for petroleum pipe lines,calculated from Construction and BuildingMaterials, Statistical Supplement, May 1954,pp. 33 and 82, and adjusted to a 1929 base.

The Commerce Department series, in 1954dollars, 1946—1963, is the 1954 value figure(Series C84), extrapolated by total footagedrilled annually.

PRIVATE IN VESTMENT IN MANUFACTURINGSTRUCTURES, COMMERCE, ANNUAL

C86. Private Purchases of Structures forManufacturing Establishments, CurrentDollars, 1929—1962.

C87. Private Purchases of Structures forManufacturing Establishments, Con-stant Dollars, 1929—1962.

SOURCE:1929—1955: U.S. Income and Output, Supple-ment to the Survey of Current Business, 1958,Table V-12, pp. 196—197. 1956—1958: Surveyof Current Business, July 1962, Table 40, p.25. 1959—1962: Survey of Current Business,July 1963, Table 40, p. 30.

For 1929—55, estimates of expenditures forstructures from the 1939 and 1947 Censusesof Manufactures and the Annual Surveys ofManufactures for 1949—53 were used as bench-marks for the privately built structures partof this series, after adjustment of the earlierfigures to cover expenditures by establish-

ments not in operation during the year, andto allow for other undercoverage. These wereinterpolated and extrapolated by Commercedata on industrial construction put in place(Series C35).

Purchases of structures built for the federalgovernment, of some significance after WorldWar II, were derived from data of the WarAssets Administration, the General ServicesAdministration, and the Federal FacilitiesCorporation. Land and pipelines wereexcluded.

The constant dollar series was derived usingthe Commerce deflator for industrial build-ings (Series C36). The figures for 1956—62 arean extension of the earlier series, using similardata and "short-cut methods." No descriptionof methods for this segment of the series hasbeen issued. This set of estimates has beendiscontinued and a revision is being preparedby the Department of Commerce.

D. Construction Materials

Dl. Manufacturers' New Orders of Construc-tion Materials, Supplies, and IntermediateProducts, Census, Monthly, 1953—1963.

D2. Manufacturers' Shipments of Construc-tion Materials, Supplies, and Interme-diate Products, Census, Monthly, 1953—1963.SOURCE:

January 1953—July 1963: Manufacturers'Shipments, Inventories and Orders: 1947—1963Revised, Series MS-i, Bureau of the Census,1963. August—December 1963: Current Indus-trial Reports, Manufacturers' Shipments,Inventories and Orders, Series MS-i, monthlyissues, October 1963 to January 1964.

These two series were initiated by theBureau of the Census in 1963, in connectionwith a major revision of the former monthlyIndustry Survey of the Office of BusinessEconomics. Data on six market categories,which cut across major industry groupingswere introduced in the new series.

The market category "construction mate-rials, supplies, and intermediate products"includes lumber and wood products, excludingwooden containers; building paper; paintsand related products; paving and roofingmaterials; all stone, clay, and glass products,excluding glass containers and kitchen articles;and fabricated metal building materials andwire products.

Data are collected from a sample panel ofmanufacturing companies. They are reportedseparately for units within a company whichare reasonably homogeneous as to type ofactivity. These units may be subsidiaries, divi-sions, departments, or plants, depending uponthe structure and activities of the particularfirm.

Monthly shipment estimates are based onthe sample, and data from the Annual' Surveyof Manufactures are used as annual bench-marks. The latter are establishment-baseddata, in contrast to the company-based datafrom Statistics of Income which were used asbenchmarks in the former Industry Survey.As an interim procedure, pending furtherstudy of the proper level for unfilled orders,a benchmark level for August 1962 was estab-lished by relating the ratio of unfilled ordersto shipments (from the sample) to the August1962 shipments estimates for each detailedcategory. The estimates for new orders arenet of cancellations. They are derived by add-ing the change in unfilled orders betweenthe current and previous month to the ship-ments estimate. Tabulation of the series onthe new benchmark basis was begun withJanuary 1960 data. Approximately compara-ble monthly data were developed for 1953—59.

Seasonal adjustment is by the source. Datashown without seasonal adjustment have beenadjusted for trading-day and calendar-monthvariation.

The source includes a detailed descriptionof the series and a comparison of the ship-ment data (Series D2) with the FederalReserve Board indexes of industrial produc-tion (Series D5).

D3. Index of Shipments of Construction Mate-rials, Federal Employment StabilizationBoard, Monthly, 1925—May 1942 (1929 =100).SOURCE:

1925—1937: Standard and Poor's Basic Statis-tics, Section 5, 1942. 1938—1942: Standard andPoor's Current Statistics, Vol. 97, No. 9, Sec-tiOn 1, September 1942.

CONSTRUCTION MATERIALS 283

This series was compiled by the FederalEmployment Stabilization Board, whose func-tions were taken over in the late thirties bythe Federal Stabilization Office, and in 1939by the National Resources Planning Board.The series was discontinued in May 1942.

By 1930, the index was composed of thirtyindividual series on the quantity of shipmentsof construction materials. Twelve series wereused beginning in 1925, and additional serieswere added as they became available. Theindividual series are weighted in the indexaccording to relative cost of the constructionmaterials.

The materials included in the index as ofJanuary 1925 were: terra cotta, floor and walltile, steel plate, enameled sanitary ware.asphalt, cement, explosives, face brick, castiron boilers, radiators, oak flooring, and mapleflooring. These were added to as follows: July1925, vitreous china plumbing; January 1926,water softening apparatus; January 1927,pumps and water systems, limestone, concretereinforcing steel; February 1927, steel boilers;September 1927, plumbing porcelain; January1928, prepared roofing, gas-fired boilers,rubber flooring, window glass; January 1929,oil burners, paint and varnish, lead pipe,oakum; January 1930, range boilers, plumb-ers' woodwork, plastic paints.

Seasonal adjustment is by the compilingagency.

For the period January 1941 to May 1942the unadjusted and seasonally adjusted dataare inconsistent as to level. Since it has notbeen possible for us to determine which iscorrect, we do not show an annual figure for1941.

An earlier index, which was a simple aver-age of six items, was compiled by AssociatedGeneral Contractors and published in theSurvey of Current Business, May 1927, p. 22,the 1932 Annual Supplement, p. 36, and laterissues. It covered 1913 through 1920, annually,and 1921 through March 1932, monthly.

D4. Index of Output of Construction Mate-rials, BDSA, Monthly, 194 7—1963 (1947—49 = 100.)SOURCE:

1947—1954: Construction Review, January1959. Revised figures for January 1952 wereobtained directly from the Office of BusinessEconomics. 1955—1958: Office of BusinessEconomics, U.S. Department of Commerce(unpublished revisions). 1959—1962: BusinessStatistics, Supplement to Survey of CurrentBusiness, 1963 Edition. 1963: Survey of CurrentBusiness, monthly issues, 1964.

The Building Materials and ConstructionIndustries Division of the Business andDefense Services Administration is responsiblefor compiling these data. The composite indexmeasures changes in the physical volume ofoutput of ten groups of construction materials,which together represented about 50 per centof the estimated value of shipments of all con-struction materials in 1947. Materials includedare iron and steel products; lumber and woodproducts; portland cement; millwork; paint,varnish, and lacquer; asphalt products; heat-ing and plumbing equipment; clay construc-tion products; gypsum products; and plumb-ing fixtures.

An index for each of the ten groups is con-structed from data on production, sales, orshipments of one or more specific materials.Data are compiled quarterly for gypsum pro-ducts and plumbing fixtures, and monthlyfor the other eight groups. Indexes are derivedon a 1947—49 base, using physical output datamultiplied by 1947 prices.

Seasonal adjustment is by the source.

D5. Index of Production of ConstructionMaterials, Federal Reserve Board, Month-

ly, 1947—1963 (1957—59 = 100).SOURCE:

1947—1960: Industrial Production, 1957—1959Base, Board of Governors of the FederalReserve System, 1962. 1961—1963: Industrial


Production Indexes, 1961—1963, Board of Gov-ernors of the Federal Reserve System, Septem-ber 1964.

Construction materials is one of the marketgroupings introduced in the 1959 revisions ofthe Federal Reserve industrial productionindexes. Indexes for most of the eighteenseries which are classified as constructionmaterials have been published monthly since1919, but the composite index begins in 1947(for a list of components and sources foreach, see Industrial Production, 1957—1 959Base, pp. S-4 to S-l9). These series, whichinclude products of both the mining and themanufacturing sectors, are from the Bureauof the Census, the Bureau of Mines, and tradeassociation sources. Some of the monthly pro-duction estimates are derived from Bureau ofLabor Statistics man-hour data, adjusted forproductivity change, but the annual indexesare all based on production or shipments data.

The index is a weighted average of rela-tives. Beginning in 1953, the weights arebased on value added, 1957—59, which isestimated by extrapolating from 1957 valueadded by the ratio of production in 1957—59to production in 1957. For 1947—52, theweights were based on 1947 value-added data;however, the weighted index was linked inJanuary 1953 to a 1957 weighted index (in the1959 revision) and subsequently converted tothe present reference period of Thecomponent monthly series are adjusted tobenchmark data from the Census of Manufac-tures and to independently derived annualindexes.

The indexes represent average productionper working day with public annual holidayscounted as working days but allowed for inthe seasonal adjustment. Seasonal adjustmentis by the source.

For a detailed description of the index, seeIndustrial Production, 1959 Revision (1959);Industrial Production, 1957—1959 Base (1962);and Industrial Production Measurement inthe United States: Concepts, Uses, and Corn-

pilation Practices (February 1964)—all publica-tions of Board of Governors of the FederalReserve System.

This series differs from Series D4 in thebase years and methods of weighting used.It also includes a wider range of products.

D6. Output of Construction Materials Des-tined for Domestic Consumption, CurrentDollars, Shaw, Annual, 1889—1939.

D7. Output of Construction Materials Des-tined for Domestic Consumption, 1913Dollars, Shaw, Annual, 1889—1 939.SOURCE: William Howard Shaw, Value

of Commodity Output Since 1869, pp. 64—65,69, and 76—77. In addition to the annual databeginning in 1889, estimates are given in thesource for 1869 and 1879 as follows: 1869($ million), current dollars, 377.4; 1913 dollars,351.4. 1879 ($ million), current dollars, 444.2;1913 dollars, 545.7.

Data represent total output of constructionmaterials, less exports and plus imports. Con-struction materials include (1) nonmanufac-tured commodities, such as lumber (destinedfor direct use in construction), crossties, sandand gravel, and crushed stone; and (2) man-ufactured products, such as lumber products,cement, brick, rails, structural ironwork, etc.(For complete list of manufactured products,see source, pp. 133—135.) These coveredcommodities include some which were classi-fled directly as construction materials, andanother group of "mixed commodities," suchas lumber, wrought pipe, and statuary andart goods, for which the proportion allocableto construction materials was estimated. Valueestimates exclude costs after production, suchas transportation and distribution costs.

For the period 1869—1919, output, exports,and imports were estimated as follows. Esti-mates of the value of output are based pri-marily on the Census of Manufactures,decennially, 1869—1899, and quinquennially,


1899—1919. Intercensal estimates were devel-oped from state statistical reports for sevenstates (see source, pp. 232—233), figures fromgovernmental agencies such as the Bureau ofMines, and the Department of Agriculture,and from trade associations such as the Amer-ican Iron and Steel Institute (see source, pp.24 1—242). Exports were compiled from detailedstatistics in the December issues of the MonthlySummary of Foreign Commerce for 1893—1919. For 1869, 1879, and 1889—1893, calen-dar-year totals were derived from the Quar-terly Report of the Chief of the Bureau ofStatistics, U.S. Treasury Department. Thesebasic export statistics were modified as toclassification of commodities, and wereadjusted to price levels comparable with theoutput data (manufacturers' or producers'prices). Imports for consumption, includingduties, were from fiscal-year data inForeign Commerce and Navigation, and wereadjusted to a calendar-year basis.

For the period 1919—33, the data are basedon Kuznets' estimates (Commodity Flow andCapital Formation, I, New York, NBER, 1938),adjusted to improve comparability with earlieryears. Ratios of Kuznets' unadjusted value ofdomestic consumption to his unadjusted valueof domestic output were computed for 1919and 1929 and interpolated for interveningyears. These ratios were applied to Kuznets'adjusted output values. After 1933 the esti-mates are rough extrapolations.

Values in 1913 prices were derived by apply-ing price indexes to the current price series.For 1914—39, Bureau of Labor Statisticsindexes for lumber and building materialsand for steel rails were combined, using 1926weights. For 1890—1913, a similar composite,using 1909 weights, was combined with anindex of structural steel prices, computedfrom Metal Statistics, 1938. For years before1890, series on eleven commodities were com-bined, using 1909 BLS weights, and the result-ing index was linked to the series for lateryears.

D8. New Orders of Rails (Tonnage) by Rail-roads, Quarterly, 1870—1926; 1924—1950.SOURCE:

1870—1926: John E. Partington, Railroad Pur-chasing and the Business Cycle, The BrookingsInstitution, 1929. 1924—1950: Compiled byNBER from Railway Age files on weeklyorders.

In the 1870—1926 segment, orders for railsrefer only to steam railroad purchases, wherepossible, and exclude streetcar and interurbantransit lines. Partington gathered his informa-tion from news items appearing in tradejournals such as Railway Age, Iron Age, andIron Trade Review. The main source for theearly years, Bulletin of the American iron andSteel Association, was used only through 1912,as weekly data were not published after thistime.

In determining the month in which ordersfor rails occurred, it was assumed that therewas as a rule an average lag of two to threeweeks between the placing of the order andthe appearance of a corresponding news itemin a periodical. Therefore, an order wasgenerally attributed to the month precedingthe one in which it was reported, unless theexact date of the order was given.

No attempt was made by Partington toadjust for possible undercoverage in the series.A comparison with data on apparent con-sumptipn (see Series D16) suggests very largeundercoverage in the earlier years and possiblyconsiderable fluctuations in coverage in lateryears (Partington, pp. 230—231). The serieswas intended for use in studying cyclicalfluctuations and is clearly unsuitable for judg-ing trends. To the extent that variations inthe ratio of reported orders to apparent con-sumption reflect year-to-year coverage fluctua-tions, rather than differences in timing betweenorders and consumption, the usefulness of theseries for cyclical analysis is also reduced.

The 1924—50 segment of the series was com-piled from rail orders reported weekly in Rail-way Age news items. For this segment, the dat-


ing of orders assumed only a three-to-twelve-day lag between the placing of an order andthe date it was published. Therefore, ordersfalling anywhere between the first and thesixth of a month were placed in the precedingmonth.

Seasonal adjustment is by NBER.

D9. New Orders of Oak Flooring (BoardFeet), Monthly, 1912—1963.

Dl0. Shipments of Oak Flooring (BoardFeet), Monthly, 1912—1963.

Dli. Production of Oak Flooring (BoardFeet), Monthly, 1912—1963.

SOURCE:1912—1922: Survey of Current Business, May1924. 1923—1933: Supplements to Survey ofCurrent Business, 1932 and 1936. 1934—1948:Lumber Industry Report, A ugust 1950 Statis-tical Supplement, Department of Commerce,Office of Industry and Commerce. 1949—1962:Business Statistics, Statistical Supplement toSurvey of Current Business, Biennial editions,1953—1963. 1963: Survey of Current Business,December 1964.

The statistics were compiled by the NationalForest Products Association (formerly NationalLumber Manufacturers Association), frommonthly reports from members and nonmem-bers, with estimates for nonreporting mills.The mills reporting represent about 75 percent of total industry output in recent years.

Beginning in 1941, the figures include pre-finished flooring. Relatively small amountsof heavy hardwood flooring other than oakhave been included since 1945. (Approxi-mately 5 per cent of the total is usually maple,beech, birch, and pecan.)


D12. New Orders of Southern Pine Lumber(Board Feet), Monthly, 1916—1963.

SOURCE:1916—1929: Southern Pine Association. 1929—1947: Lumber Industry Report, StatisticalSupplement, U.S. Department of Commerce,

Office of Industry and Commerce, August1950. 1948: Office of Business Economics, U.S.Department of Commerce (unpublishedrevisions). 1949—1960: Business Statistics,supplement to Survey of Current Business,Biennial Editions, 1951 through 1963; and(for 1954 data) Survey of Current Business,November 1957, p. 24, U.S. Department ofCommerce. 1961—1962: Office of BusinessEconomics, U.S. Department of Commerce(unpublished revisions). 1963: Survey ofCurrent Business, February and September1964.

Data were compiled by the National ForestProducts Association (formerly National Lum-ber Manufacturers Association), from monthlyreports from individual mills. Figures for the1916—1929 segment cover 192 mills; those forthe 1929—63 segment are estimates for all pro-ducers of Southern yellow pine, based on datafrom reporting mills accounting for approxi-mately one-sixth of total output in recentyears.

Data are adjusted to the level of the annualproduction figures published by the Bureauof the Census (except for 1948, when no Cen-sus survey was made). Since there was con-siderable undercoverage in the Census dataprior to 1942, data for those years are under-stated to an unknown degree. <


D13. New Orders of Fabricated StructuralSteel (Tonnage), Mont lily, 1909—1963.

SOURCE:1909—1922: Record Book of Business Statistics,Part II, U.S. Department of Commerce, 1927.1923—1932: Survey of Current Business, 1932Supplement and issue of March 1933. 1933—1963: American Institute of Steel Construc-tion, Inc. Current data are available in Sur-vey of Current Business.

Data represent estimates for the total fabri-cated structural steel industry. They cover


only structural steel fabricated to order foruse in construction. Products such as windowand door frames, stairs, and ornamental ironare not included. Beginning in 1952 certainitems not previously covered were included(intercompany and export work). New ordersare net and pertain to actual contracts closed.

Estimates for 1909—32 were prepared by theBureau of the Census, based in part on datacompiled by the Bridge Builders and Struc-tural Society and the Central Fabricators'Association. Percentages of capacity for cov-ered firms were applied to total capacity toderive total industry estimates. For 1933—42and 1946—63, estimates for the entire industrywere prepared by the American Institute ofSteel Construction, based on reports frommembers who account for over 50 per centof total industry shipments. For the years1943—45, the Institute compiled data only ontonnage actually reported. Industry totalsshown for these years were derived by NBERon the assumption that coverage was 80 percent.

The latest revision of the series by theA.I.S.C., to the level of the 1958 Census ofManufactures, covers only the years from1947 on. We have therefore shown an overlapin 1947.


D14. Index of Lumber Production, FederalReserve Board, Monthly, 1919—1947(1935—39 = 100); and 1947—1963 (1957—59 = 100).

SOURCE:

1919—1942: Federal Reserve Index of Indus-trial Production, Board of Governors of theFederal Reserve System, October 1943. 1943—1947: Federal Reserve Bulletin, Board of Gov-ernors of the Federal Reserve System, monthlyissues, 1944—48. 1947—1960: Industrial Produc-tion, 1957—59 Base, Board of Governors of theFederal Reserve System, 1962. 1961—1963:Industrial Production Indexes, 1961—1963,Board of Governors of the Federal ReserveSystem, September 1964.

This index measures changes in the physicalvolume of lumber production, and does notcover logging and lumber products (miliwork,etc.). The basis for the index is the NationalLumber Manufacturers Association series onlumber production, in board feet. This seriesis currently available in the Survey of CurrentBusiness, and a brief description of it appearsin Business Statistics, 1963 edition.

This index is a component of the FederalReserve Index of Production of ConstructionMaterials, 1947—1963 (Series D5). The indexesfor individual series are available on the1957—59 base only from 1947; for earlier yearsthe latest data are on a 1935—39 base. Thisseries is therefore shown here in two segmentswith an overlap of one year.

For a summary of the methods used in cal-culating the indexes, see notes to Series D5.Lumber production constituted about 10 percent of total output of products classified asconstruction materials in 1957—59.

Seasonal adjustment for 1919—47 is by thesource; and for 1947—63 by NBER. The orig-inal and seasonally adjusted data have beenadjusted for variations in the number of work-ing days.

For a detailed description, see above sources,references given in Series D5, and FederalReserve Bulletin, December 1953.

D15. Index of Production of Clay, Glass, andStone Products, Federal Reserve Board,Monthly, 1919—1947 (1935—39 = 100);and 1947—1963 (1957—59 = 100).

SOURCE: Same as that for Series Dl4.

This index measures changes in the physicalvolume of production of clay, glass, and stonemanufactured products. It does. not coverstone and earth minerals. It is based on pro-duction or shipments data from the Bureauof the Census, the Bureau of Mines, and tradeassociations, such as the American GlasswareAssociation; and on BLS man-hours data.

The index is available on the 1957—59 baseonly from 1947; for earlier years the latest


data are on a 1935—39 base. This series istherefore shown here in two segments with anoverlap of one year.

For a summary of the methods used in cal-culating the index, see notes to Series D5.Most clay, glass, and stone products (85 percent of output in 1957—59) are classified asconstruction materials and are included inthe index for that market group. In 1957—59they accounted for one-quarter of totalproduction of construction materials.

Major revisions of the clay, glass, and stoneindex were made in 1940, 1943, 1953, 1959,and 1962. In addition to changes in the baseperiod, the revisions affected the compositionof the index, weighting, adjustment to bench-mark levels, types of basic data used, andmethod of seasonal adjustment.

Seasonal and working-day adjustments areby the source.

For a detailed description of the index andthe revisions, see above sources, referencesgiven in Series D5, and Federal Reserve Bulle-tin, December 1958.

D16. Rail Consumption (Tonnage), Annual,1849—1961.

SOURCE:

1849—1940: American Iron & Steel Institute,Annual Statistical Reports. 1941—1961: MetalStatistics, American Metal Market, 1962 and1963.

Rail consumption estimates are computedas rail production plus imports and minusexports. Prior to 1872 figures are only forproduction plus imports. Data through 1940were computed by the American Iron andSteel Institute. Data beginning 1941 fromMetal Statistics are based on AISI productionfigures, and import and export data are fromthe U.S. Department of Commerce.

The data cover total production of ironand steel rails. They include rerolled rails,girder and high tee rails, and streetcar rails.

D17. Portland Cement Production, Monthly,1911—1963.

D18. Portland Cement Shipments, Monthly,1911—1963.

SOURCE:191 1—1920: George Otis Smith, et al., "Fluctua-tions in Mineral Output," The Problem ofBusiness Forecasting, papers presented at theEighty-Fifth Annual Meeting of the AmericanStatistical Association, edited by Warren M.Persons, William Trufant Foster, and AlbertJ. Hettinger, Jr., Cambridge, 1924. 1921—1962:Survey of Current Business, September 1923,and Supplements, 1932—63. 1963: Survey ofCurrent Business, monthly issues, March 1964to February 1965.

The data relate to finished portland cement,measured in number of barrels, for the entireindustry. They include high-early-strengthcement which, beginning 1955, is separatelyreported by the compiling agency. Clinkercement, an intermediate product, is excluded.

The 191 1—1920 data were compiled by E. F.Burchard of the U.S. Geological Survey, fromdata furnished by the Portland Cement Asso-ciation. The original data, representing 93 to99 per cent of industry production, weresupplemented with estimates for nonreportingplants.

Data after 1920 are from the Bureau ofMines, U.S. Department of the Interior. Theyinclude 48 states and the District of Columbia;Puerto Rico (where operation was started in1940); and Hawaii (for September 1944through 1946 and beginning 1961).


D19. Asphalt Production, Monthly, 1917—1950(Short Tons), 1951—1963 (Barrels).

SOURCE:1917—1930: Petroleum Refinery Statistics, U.S.Bureau of Mines, Bulletins 280, 289, 297, 318,339, 367. 1931—1940: Minerals Year Book, U.S.Bureau of Mines, 1931—1940. 1941—1962: Sur-


vey of Current Business, Biennial Supple-ments, 1947—1963. 1963: Survey of CurrentBusiness, monthly issues, February 1964 toFebruary 1965.

The Bureau of Mines is the official sourcefor these data. Figures through 1958 are forthe U.S., excluding Alaska, Hawaii, and PuertoRico. Alaska and Hawaii were includedbeginning January 1959. The data cover only

asphalt made from petroleum, domestic andimported.

Monthly data for years through 1950 werereported in thousands of short tons; beginningin 1951, data were reported in thousands ofbarrels. We have converted the earlier annualfigures to barrels, on the basis of 5.5 barrelsto a short ton. Monthly data presented hereare as reported in the source.


Documents

PDF (1036 K) - National Bureau of Economic Research