For directions on reading the variable description see Data Dictionary Introduction.

Availability: 
 

1850	1860	1870	1880	1900	1910	1920	1940	1950	1960	1970	1980	1990
X	X	X	X	X	X	X	X	X	X	X	X	X

Universe: 
All persons. 

Codes and Frequencies - General, Detailed

Description: 
BPL indicates the person's state or foreign country of birth. 

Comparability: 
While this variable yields similar information for all years, each census had to grapple with shifting political boundaries and an increasingly diverse population that required more inclusive classification schemes in later years. The ways in which these and other factors affected (and limited) our efforts to make BPL comparable across years are described below. 

In general, the BPL general code describes places available in multiple years. The first digit of BPL more or less corresponds to continents. The detailed code notes places not available over a number of years, or places that are strictly subsets of other specified countries. Many detailed codes relate to 1980 and 1990 codes; years in which the level of geographic detail reached new heights. 

Users might also examine other IPUMS variables dealing with birthplaces, ancestry, and language to determine which variables are most useful for their purposes (MBPL, FBPL, MTONGUE, MMTONGUE, FMTONGUE). 
Factors affecting persons born in the United States: 

1. Boundaries of some U.S. territories and states changed over the 140 years covered by the IPUMS; see maps in "Geographic Variables and Tools," Volume 2: User's Guide Supplement. Beginning in 1900, the censuses tried to deal with this problem in one way or another, but it is not always clear how accurately enumerators and respondents were able and/or willing to follow census instructions.

   Prior to 1900, the census provided no explicit instructions to enumerators other than to report the state or territory in which the person was born. From 1900 onwards, native-born Americans were to follow contemporary boundaries when reporting the state, territory, or U.S. possession in which they were born. 
2. The 1950 enumerator instructions and the 1960, 1970, and 1980 census forms further instructed respondents to give their state (or country) of birth as the state (or country) in which their mother resided when they were born, even if their mother went to a hospital in another state (or country) to give birth. This instruction was dropped in 1990 when census evaluations showed that many respondents either ignored or misunderstood it.

1. Boundaries of several foreign countries changed over the 140 years covered by the IPUMS, as did the detail with which the census instructed enumerators and/or respondents to report birthplaces. While we can explain how each census dealt with this problem in general, users must be aware of particular shifts affecting the immigrant groups they study. Central and Eastern Europe around the time of the World Wars poses some special difficulties. For some countries and years, the census gave specific instructions, but for others enumerators and respondents were free to report their birthplaces as they saw fit. Furthermore, it is not always clear how well enumerators and/or respondents followed the census's instructions.

The makers of each sample usually only preserved as much detail as they thought would be useful to potential users. While censuses usually recorded country of birth, some samples (especially the earlier ones) combined groups of non-European countries-such as those of Central America, South America, or Africa-into one category, since the number of cases for each individual country was thought to be too small to be useful. In some PUMS detail below the nation-state level is available. Users can examine the codes and frequencies table to determine the amount of detail preserved in each sample (and thus in the IPUMS). 
2. Beginning with the 1900 census, persons born abroad to American parents were to indicate this, and were not to give a separate country of birth. These cases are coded "900" under the general category "Abroad."
3. Persons born "at sea" were instructed to indicate this status beginning with the 1900 census. Some people did so in previous years, and they are so coded in the sample and IPUMS.

1850:

Enumerators simply recorded whatever birthplace the respondents reported. Some reported countries or regions (such as Poland) that did not exist as nation-states at the time of enumeration; this occurred frequently in later years as well.

1860-1880:

Instructions were not much more explicit than those for 1850. Enumerators were to record the country "as specifically as possible." Those born within Great Britain were to specify England, Wales, Scotland, or Ireland. Those born within Germany were to specify the particular state. The census did not issue more elaborate instructions to enumerators. As in 1850, the detail of responses varied and did not always accord with contemporary political boundaries. To cite only one example, most people born in Canada simply replied "Canada," but more than 1000 persons included in the sample responded with a specific Canadian province (New Brunswick, Nova Scotia, Ontario, Quebec, etc.).

1900:

Enumerators were instructed to record the nation-state of the person's birth-that is "a region whose people have direct relation with other countries." The instructions made several exceptions to this general rule: as in 1880, enumerators were to distinguish the various parts of Great Britain. They were also to identify three areas separately-Austria, Hungary, and Bohemia-within the Austro-Hungarian Empire, and they were to separately identify persons born in Finland (ruled by Russia in 1900). They were also to determine and record whether people responding "Canada" were of English or French descent. Persons who initially responded "Poland" were to specify Austrian Poland, German Poland, or Russian Poland. In all of these cases, most respondents apparently complied, but the 1900 PUMS includes a few who responded "Austria-Hungary" or "Poland"-these either could not or would not give a more specific answer. Furthermore, nationalism might have played a role in some respondents' answers; for example, a person of Russian descent who was born in Finland may have responded "Russia," even though the instructions intended otherwise.

1910:

Enumerators were to distinguish between the various parts of Great Britain and to separate Finland from Russia, as in 1880 and 1900. They were also to separate Austria and Hungary-but unlike the 1900 instructions, those for 1910 made no mention of Bohemia. 1910 also differed from 1900 in that no mention was made of the areas of Poland or the French-English Canadian distinction, while it asked enumerators to make a new distinction between Turkey in Europe and Turkey in Asia. Users who examine the 1910 frequencies will find that, as in previous years, some enumerators or respondents gave answers without heed to these instructions.

1920:

Enumerators were again to distinguish between the various parts of Great Britain. They were additionally instructed to distinguish Cuba and Puerto Rico from the rest of the West Indies. In order to account for the impact of World War I on the political geography of Central and Eastern Europe, enumerators were to ask those born in Austria-Hungary, Germany, Russia, or Turkey, as defined by their prewar boundaries, to specify the province, state, or region they were born in.

1940:

Enumerators were to record the country in which the respondent's birthplace was situated as of January 1, 1937 or, if this was not known, they were to record the province, state, or city, in which case editors later fit them into the appropriate 1937 nation-state boundaries. Note that this was the first of the censuses contained in the IPUMS to explicitly instruct enumerators to determine the country in which each birthplace was located at a point in time relatively contemporary with the census-in this case they used 1937 instead of 1940 to get around the political shifts occurring just before and during World War II. This may have been the intention prior to 1940, since birthplaces within the United States had long been recorded according to contemporary state boundaries and place names (see "Factors affecting persons born in the United States," above). Still, the census gave no explicit instructions in this regard for recording foreign birthplaces, except for those applying to specific nation-states, which are described for each year above.

The 1940 census continued to separately identify the various provinces of Great Britain. Ireland was now independent and distinguished from the UK's Northern Ireland. Cuba and Puerto Rico (as well as the Virgin Islands) were now to be distinguished from the rest of the West Indies, and the census re-instituted the 1900 distinction between French and English Canada (as well as Newfoundland) and the 1910 distinction between Turkey in Europe and Turkey in Asia. 

The creators of the 1940 sample, did not precisely follow the 1940 census coding scheme. They did not distinguish between the two parts of Turkey, and they did not keep Danzig or Newfoundland as separate categories (the 1940 census did). As in previous years, some enumerators reported areas that were not actually independent states in 1937, and the sample preserved these responses; users should check the codes and frequencies table carefully to identify such categories. 

1950:

Enumerators were to record the full name of the foreign country according to contemporary (1950) international boundaries, or, if this was not known, they were to record the locality as precisely as possible, in which case editors later fit them into the appropriate country.

The 1950 census made exceptions to this rule similar to those made in 1940: the provinces of Great Britain were separately identified, the Ireland/North Ireland distinction maintained, and French Canada distinguished from the rest of Canada. For all persons born in the West Indies, the precise island was to be ascertained. As in previous years, some enumerators reported areas that were not actually independent states in 1950, and the PUMS preserved these; users should check the codes and frequencies table carefully to identify such categories. 

1960, 1970, 1980, and 1990:

Respondents were to report their place of birth according to contemporary international boundaries as recognized by the United States. Some subregions, such as the British provinces and some parts of the Soviet Union, are identified separately in the sample-see the codes and frequencies table. The 1960 and 1970 samples continued the practice of combining some (mostly non-European) countries into larger sub-continental regions when the numbers of people from each country within the region was small. The 1980 and 1990 samples gave a separate code to each country (and some smaller subregions).

The most time-consuming task was to determine how each country contained in one sample coding scheme, but not in another, would have been classified in the scheme that did not contain it. Most of these cases were fairly straightforward; usually, they were countries that were identified separately in some samples but grouped with other countries into a more inclusive category in other samples. For example, Argentina (IPUMS code 300-05) was separately identified in all samples except for those for 1900 and 1910, in which all South American countries were collapsed into the category "South America."  In such cases, all countries contained in the more inclusive category are given the same IPUMS general code- "300" in the case of South America-and the detail contained in other years is preserved in the detail codes (the second two digits). (Of course, some responses rarely or never occurred in the earlier censuses because few or no people from a particular country had entered the United States.) 

In cases involving changes in political geography, the IPUMS coding scheme in effect imposes its own artificially static political boundaries. In some cases, countries (such as Korea) have broken into two or more independent units, or they have been combined (like the German states or, later, East and West Germany) to create one larger unit. Generally, the IPUMS gives the largest unit ("Korea" or "Germany") a general code, while the smaller units are given separate detail codes for the years in which they existed and are available. Thus, Korea has the IPUMS general code 506; the detailed codes are 506-00 for "Korea," 506-10 for "North Korea," and 506-20 for "South Korea." 

Some countries, such as France and Germany, have changed in size over the years covered by the IPUMS. While some of the census instructions described above tried to deal with this problem-and their success always depended upon how willing or able respondents were to reply correctly-the IPUMS can do little about it. The response "Germany" is given the same code (453-00) in all years, even though Germany's boundaries have shifted; the census instructions might also have affected the number of people who gave "Germany" as a response. 

The various "other" categories can not be made compatible across years. The content of these categories varies with the list of specified places in the given region. 

The IPUMS staff has no comment on the Antarctican population identified in the censuses of 1980 and 1990. 

Flags: QBPL 

Census Questions:
1850   1860   1870   1880   1900   1910   1920   1940   1950   1960   1970   1980   1990

Enumerator Instructions:
1850   1860   1870   1880   1900   1910   1920   1940   1950   1960   1970   1980   1990