Relation of the Federal GoveTnment to Research 



209 



which are of general public interest, even though such action 

 involves a duplication of effort."' 



It is cleai' that however vahiable the Bureau's cur- 

 rent industrial reporting might be to economists, or 

 to a Government agency, which, for example, wanted 

 unbiased data for use in arbitrating a labor dispute, 

 this work is not necessarily an unmixed blessing to tlie 

 Bureau from the standpoint of winning general sup- 

 port from industry. 



Need of a field staff. — An old dream of friends of 

 the Bureau of the Census which has not materialized 

 is the possession of a permanent field staff, with re- 

 gional headquartei'S in important cities. If each head- 

 quarters could be in charge of a full-time man who 

 not only understood the technical statistical problems 

 of the Bureau of the Census, but also had the dynamic 

 qualifications of a good sales manager, and the per- 

 sonality to win the respect of trade associations, cham- 

 bers of commerce, and individual manufacturers 

 through public addresses and private conferences, the 

 problem of winning the cooperation of industry for 

 the whole Bureau program might be on the road to 

 solution. Moreover, such a field office could be a head- 

 quarters not only for the Census of Manufactures and 

 Census of Business but also for the Census of Popula- 

 tion. Instead of temporary supervisors or low-paid 

 career clerks temiDorarily dispatched from the Wash- 

 ington office, the Bureau of the Census would have 

 permanent representatives who would get the worlc 

 done and build prestige. 



The Bureau has had ample opportunity to observe 

 the importance of a permanent field staff. An example 

 may be cited. The Census of Mines and Quarries is 

 the responsibility of the Bureau of the Census. 

 Since the Bureau of Mines has field representatives 

 throughout the mining areas, the responsibility for 

 the collection of the data was turned over entirely to 

 this staff at the last Census. These men understood 

 such difficult technical problems as determing whether 

 a given mine was a copper, or silver, or zinc mine 

 and they knew intimately most of the establishments 

 from whom schedules were to be secured. 



Rules of secrecy. — If closer continuous contacts 

 could be established with industry by the Bureau of 

 the Census, it might be possible eventually to relax 

 to a greater extent the rules of secrecy, which guarantee 

 that no information about an individual firm will be 

 revealed. The presentation of only fragmentary data 

 for States and cities in the published tables of the 

 Census of Manufactures whenever a full report would 

 reveal facts about an individual plant is a large price 



'^Report of Committee on Government Statistics, pp. 32-33. There 

 are several Committee memoranda on the subject, listed In Appendix C 

 ot the Committee's report. 



to pay for cooperation. The absurd totals which ap- 

 pear for many States and cities exasperate almost 

 everybody who tries to put the data to scientific use. 

 The Committee on Government Statistics felt so 

 strongly on this point that it recommended that com- 

 plete figures on at least the total number of wage earn- 

 ers be published for all geographic areas and that 

 special legislation be passed, if necessary, to modify 

 the secrecy rule.=* So much suspicion on the part of 

 manufacturers is encountered, however, lest their cen- 

 sus reports fall into the hands of the Federal Trade 

 Commission or the Bureau of Internal Revenue, that 

 the Bureau of the Census is leaning over backward to 

 protect respondents. Otherwise, the Bureau fears that 

 it could not get responses without resort to legal pres- 

 sure or else that the statistics would be "doctored." 

 Every Bureau letterhead carries in red ink the words 

 "Your Census Reports are Confidential" and the cam- 

 paign to win confidence and secure accurate replies is 

 perhaps more active today than ever before. 



Federal-State Relationships in 

 Vital Statistics 



Frequently one hears suggestions that the statistical 

 work of the Federal Government be decentralized and 

 a larger proportion of the task be performed by the 

 individual States. The experience of the Bureau of 

 the Census in connection with vital statistics — that is, 

 statistics of births and deaths — -provides an illuminat- 

 ing illustration of the type of problems encountered. 



The State boards of health collect the vital statistics 

 and they are also among the foremost consumers of 

 these statistics, using them in various j^ublic health 

 programs. The Division of Vital Statistics in the 

 Bureau of the Census compiles national figures based 

 on the birth and death certificates which are collected 

 by the States and transcribed by the States for trans- 

 mission to Washington. The Bureau of the Census sub- 

 sidizes the States in their work. 



The Federal Government, for reasons discussed in 

 Part I of this report, lagged far behind most of the 

 European countries in developing current national re- 

 porting of births and deaths. A strong motivation 

 for establishing a permanent Bureau of the Census was 

 the recognition of the need, by public health officials, 

 actuaries, and students of population, of better national 

 vital statistics than were obtained once in 10 years in 

 the house-to-house canvass for the Census of Popula- 

 tion. 



Beginning as of 1900, the Bureau adopted the plan 

 of establishing registration areas for births and deaths. 

 States would be admitted which passed uniform legis- 



'* Report of Committee on Oovernment Statistics, p. 48. 



