NARA expects to digitally release the 1950 population census schedules for researcher use on April 1, 2022, which is 72 years after the official 1950 census.  This is the 11th in a series of blog posts on the 1950 census.


During the 1950 census, the Bureau of the Census gave special attention given to infants.  For each infant born in January, February, or March 1950, who was listed on the Form P1, Population and Housing Schedule, the enumerator was required to take the additional step of filling out Form P3, Infant Card.  Although most of the information on the Infant Card could be copied from the population schedule, the enumerator had to ask the family for additional information, such as actual place of birth, name of hospital, type of attendant, and maiden name of mother.


Advertising image for 1950 Census from [Folder 13] Advertising Campaign for the 1950 United States Census, 17th Decennial Census (National Archives Identifier 195980247).


Form P3, Infant Card, measured 8 inches by 10.5 inches. Three million cards were printed in green ink on buff colored paper.  Blank Infant Cards were distributed to census enumerators in the Portfolio that contained their enumeration district description, maps, blank forms, instructions, and other needed supplies.  Enumerators in certain U.S. territories and possessions also filled out Infant Cards that were substantially the same as the P3, Infant Card, but were numbered differently:  P84, Alaska; P89, Hawaii; P95, Puerto Rico; and P99, Virgin Islands.  This is an image of the Form P3:


Form P3, Infant Card, from [Folder 12] Form P3, Infant Card, 1950 Census of Population and Housing, 17th Decennial Census, 1950 (National Archives Identifier 195980246).


The Infant Cards served as the basis for studying the “undercount” of babies in two ways.  The cards were to be matched by the National Office of Vital Statistics to birth registrations to determine (1) how many enumerated infants were not registered at birth and, conversely, (2) how many infants registered at birth were not enumerated in the 1950 census. Sadly, many infants born during January to March 1950 could not be enumerated on the census because they died before the official census day of April 1, 1950.  In 1950, nationwide infant mortality was 29.2 deaths per 1,000 live births, with white infant mortality at 26.8 per thousand and Black infant mortality substantially higher at 43.9 per thousand.


After the census was completed, each Infant Card was removed from the Portfolio for its enumeration district and then grouped by state and finally by county or large city.  Occupation and industry entries for the fathers were coded (tabulated) for statistical tabulation purposes “before the National Office of Vital Statistics matched the cards against the birth registration records.” In addition to the Infant Cards, some data was obtained from a special mail inquiry “to parents of infants for a sample of birth records for which no matching census record could at first be found.” Possible reasons for missing of infants in the census were also obtained through the same mail inquiry. The Bureau then proceeded to study the data. As the Bureau stated:


“The Infant Enumeration Study was undertaken to provide information that might lead to better enumeration of young children in future censuses and to provide a basis for estimating the extent of error in 1950. Although it was necessary to restrict the study to infants born in the first three months of 1950, some of the patterns of differential completeness of enumeration by urban-rural residence, etc., may be applicable to older infants.”  1950 Infant Enumeration Study, p. 1.


The full study is available online and is worth reading.  Here are the highlights from the introduction (page 1):


  • About 96.4 percent of infants born in the first three months of 1950 were enumerated in the 1950 Census. This figure includes 1.3 percent for whom birth records indicated that the infants were born in the period but were enumerated as born before 1950. This figure excludes infants whose birth records indicated that they were illegitimate or were born in a State other than the State of usual residence of the mother.
  • Taking into account illegitimate infants and infants born in a State that was not the usual residence of the mother, 95.1 percent of infants born in the first three months of 1950 were enumerated in the 1950 census.
  • The North Central Region had the highest percentage of infants enumerated (97.8 percent) and the South the lowest (94.7 percent).
  • Ohio and Connecticut led with 98.3 percent and New Mexico was last with 91.0 percent. Thirty-four of the States had percentages of 95.0 or better.
  • Enumeration of infants varied slightly by where they lived: urban (96.8 percent), rural nonfarm (96.7 percent), and rural farm (94.7 percent).
  • White infants (97.1 percent) were more completely enumerated than nonwhite infants (91.4 percent).
  • There was no evidence of any real difference in the enumeration of boys and girls.
  • By month of birth, there was slightly less complete enumeration of babies born in January than those born in February or March.
  • By order of birth, there was less complete enumeration at the extremes. "First child" and tenth-or-higher order babies were less completely enumerated than the intermediate birth orders. These comparisons may reflect unsettled living arrangements among some recent parents of a first child and among some very large families.
  • Infants of mothers under 25 years old tended to be less completely enumerated than infants of older mothers.
  • Infants of mothers who had completed less than 7 years of grade school were more likely to be missed than infants of mothers with more education.
  • Infants with fathers who were in the armed forces were relatively more often missed than were infants with civilian fathers.
  • About 82 percent of the 16,045 cases in which infants were classified as definitely or probably missed in the census, the parents were also missed.  Of these:

o   20 percent of the families of missed infants were absent from home or moved during the enumeration period.

o   55 percent of the missed infants were not counted because the enumerators overlooked some obscure dwellings, failed to enumerate all dwelling units in a structure, or listed some occupied dwelling units as “vacant.”

o   8 percent resulted from failure of relatives or nonrelatives to report the parents and the infant who were staying with them, probably because the enumerator did not ask specifically about people who were living temporarily in the dwelling unit.

  • In the 18 percent of cases where the parents were enumerated but the infant was missed, the reasons given for the oversight included:

o   A neighbor gave incomplete information.

o   The family did not think infants were to be reported or forgot.

o   The infant died between April l and the time of the census enumeration.

o  The child was born before April 1 but was not enumerated because the enumeration occurred before April 1 in an enumeration district used for training enumerators. (1950 Infant Enumeration Study, p. 8). 


Form P3, Infant Cards, will not be part of the release of the 1950 census.  These cards are not extant.  They are not part of the holdings of the National Archives and Records Administration (NARA). The writer has not researched the details. However, the Infant Cards were undoubtedly deemed to be temporary records that were not worthy of permanent retention because they (1) duplicated state birth records kept by state vital records agencies and (2) largely duplicated data already included in the permanent record in microfilm publication T628, Seventeenth Census of the United States, 1950, that contained the population side of the Form P1, Population and Housing Schedule.  The Bureau of the Census microfilmed the 1950 census population schedules in 1952 prior to disposal of the original paper schedules.  The Bureau's microfilm was then accessioned by the National Archives.


While the Infant Cards are an interesting aspect of the 1950 census, it is worthwhile noting that Infant Cards were not a new or innovative feature of U.S. census taking.  The Bureau of the Census had, in fact, also required enumerators to complete Infant Cards in conjunction with the 1940 census for infants under 4 months old.  The Bureau estimated that the count of children under one year old may have been less complete in 1940 (despite the Infant Cards) than in 1930 (when there were no Infant Cards).  1950 Infant Enumeration Study, 1950, p. 3. Infant Cards for 1940 are also not extant.


Overall, the Bureau estimated that the net under-enumeration of the population of all ages in the 1950 census was fairly low - just 1.4 percent.  1950 Infant Enumeration Study, p. 3.




Department of Commerce, Bureau of the Census. Procedural Studies of the 1950 Census No. 1:  Infant Enumeration Study (Washington, DC: 1953).


Department of Commerce, Bureau of the Census. Procedural Studies of the 1950 Census No. 2:  The 1950 Censuses—How They Were Taken (Washington, DC: 1955), pages 20, 27, 28, 85, 87, 93, 190, 199, 211, 220.


Department of Commerce, Bureau of the Census.  “Status of the 1950 Census Program in the United States." A special report to the Fourth Session of the Committee on the 1950 Census of the Americas, Inter-American Statistical Institute, Washington, DC, June 11-15, 1951.


Centers for Disease Control:  Table 11, Infant mortality rates, by race: United States, selected years, 1950-2015.


National Research Council Committee on National Statistics, Vital Statistics:  Summary of a Workshop (Washington, DC: 2009), Appendix B:  The U.S. Vital Statistics System: A National Perspective.  In 1950, the National Office of Vital Statistics was part of the Federal Security Administration; its functions are currently under the umbrella of the Centers for Disease Control and Prevention (CDC) in the U.S. Department of Health and Human Services (HHS).