28 October 2000
Minutes of Meeting held at the Hilton Towers, Pittsburgh (7.30-9.00am)
Those present: Peter Baskerville, Gordon Darroch, Lisa Dillon, Soren Edvinsson, Chad Gaffield, Todd Gardner, Olof Gardarsdottir, Patt Kelly-Hall, Bob McCaa, Kees Mandemakers, Steve Ruggles, Matt Sobek, Gunnar Thorvaldsen, Matthew Woollard.
1. Minutes of previous meeting (9 August 2000)
The minutes of the previous meeting were still in the process of editing. Copies had been pre-circulated.
2 Update of members’ activities.
2.1 Norway (Gunnar Thorvaldsen)
- Gunnar now had complete 1801 census from Jan Oldervoll, University of Bergen.
- The samples for 1865 and 1900 are now complete, and on the web, but the user interface is crude, and beta-testing on a new system is in process. Note: there can be no sub-setting by ethnicity for 1900.
- A start has been made on the 1875 census. Approx. one-third will be ready by November. Coding will start in the new year. Hopeful that part will be ready in IPUMS format shortly.
- Statistics Norway will be publishing original documentation for a variety of census as part of their 125th anniversary, including a $10,000 grant to Gunnar, but has little control over content.
- Notes problems with Norwegian occupations being coded into HISCO.
2.2 Sweden (Soren Edvinsson)
- Still progressing from last meeting. Parish statistics will hopefully be available early 2000.
- Plans afoot for a new sample of the Suni population [check spelling].
2.3 UK (Matthew Woollard)
- 1881 100% sample progressing. A 2% sample to match the Michael Anderson 1851 2% sample was being created, though some difficulties with Scottish geography.
- Possibilities regarding 1901 100% sample, which is being created by Public Record Office. A machine-readable index was being created to facilitate genealogists' (and others) searching the digitized enumerators’ books. There is a possibility that this index might be made available to the academic community, but given that this is a commercially sensitive project this was not yet in the bag.
- 1891-1921 samples created by ONS for Cambridge Group had been “withdrawn”, surrounding confidentiality issues in the run up to 2001 census. Bob McCaa suggested talking with Chris Denham.
[At this stage there was a general discussion about confidentiality in censuses, and identifying people in the past. Steve mentioned that the only time that this issue had arisen to his knowledge was surrounding a divorce case, but this was using survey data. Gunnar mentioned that in Norway, the fact that German’s confiscated the radios of Jews in Norway during the second world war pointed to “misuse” of these records. It was noted that some more recent censuses were open, e.g., 1945 Newfoundland, 1960 Iceland and 1940 Family Cards from the Netherlands. It was suggested that a) we might try and find out and circulate to each other, other instances of where historic census data may have caused mishaps and b) that we don’t publicise that some recent censuses are open as an argument for more openness, as this might lead to the closure of these more recent censuses. Ruggles mentioned the Cato Institute as having a profound difficulty in scaremongering on this issue, but Gardner suggested that the US Census Office were generally following this old line, in the absence of rational argument. Ruggles mentioned the 2000 PUMS. Further discussion on the role of EU for European records followed, mentioning the conflicting laws on privacy and publications, and on the Data Protection Act. See, for example http://www.pro.gov.uk/recordsmanagement/access/access1.pdf.
2.4 Netherlands (Kees Mandermakers)
- Data entry is progressing
- The protocol promised in the last minutes has not been circulated yet.
2.5 Iceland (Olof Gardarsdottir)
- Olof and Gunnar had met with National Archives of Iceland regarding making some of the Icelandic census data available on the web.
- The Icelandic situation is complicated by the activities of DCODE, a private company researching genetics in the Icelandic population. The Archives are willing participants in the project, but DCODE are less willing. There are numerous databases, of variable quality of Icelandic censuses, notably, 1703, 1801, 1816, 1835, 1845, 1870, 1901 and 1930.
2.6 USA (Steve Ruggles/Bob McCaa)
- IPUMS collaborating with Population Studies Institute at Michigan to develop DDI tools for the census data. The resulting software, ready in a couple of years?, will be freely available.
- IPUMSi will be moving to a meta-data driven extraction system
- IPUMSi are scanning all unpublished statistical material from 1960s from almost all countries. (This is mainly the enumeration documentation, but other materials too. (From UN Statistics Service Archive.)).
- Kenya 5% samples for 1989 and 1999 deposited with IPUMSi. And Austria are having discussions with Bob towards making a sample for 1961 available.
2.7 Canada (Peter Baskerville/Lisa Dillon)
- End of project conference for Canadian Family History Project to be held in Toronto in April 2001. Data will be publicly released.
- Some funding problems surrounding the 1881 Mormon-created database, but the Latter-day Saints and University of Ottawa had provided some money for cleaning.
- Kris Inwood was attempting to get funding for a new sample from the 1891 census.
3 Conference planning
3.1 SSHA 2001
- Diana Magnuson had suggested a session at SSHA 2001 on census enumeration procedures for the 1880/1 Triangle project.
- Ruggles suggested that some early results might be aired at this conference. (Any specific areas?)
- Gaffield suggested that we might consider a session based on “why the 100% samples are better than the 2% samples”.
- Dillon suggested a session on new approaches in comparative history as a special issue of Historical Methods.
3.2 Confidentiality consultation
- Gaffield will circulate a proposal for a specific IMAG workshop on confidentiality issues.
3.3 Other meetings
- A further meeting was proposed for the AHC Conference in Poznan, Poland, but was deemed unlikely because of the uncertainty of that conference. However, it was suggested that the European end of IMAG might propose a general session to create awareness of IMAG and on census matters generally.
4 Next IMAG meeting.
Two proposals. SSHA Chicago (October 2001) or as an adjunct to the CFP conference in Toronto in April 2001. Depending on developments a decision will be made.
2.4 Mandermakers to circulate protocol on best practice for large databases.
Gaffield to circulate a proposal for a series of notes on current issues in census as a special issue/section of Social History/Histoire Sociale.
3.1 Dillon to get Diana Magnusson to circulate a proposal for SSHA 2000 on the IMAG list.
3.1 Dillon to circulate proposal for session on comparative history session on IMAG list.
3.2 Gaffield to circulate a
proposal for a session on confidentiality (where?, when?)