FOI for government statistics

I look forward to a Freedom of Information Act that will, in particular, increase public access to the raw data gathered in the government’s censuses and surveys, which underlie its published summary statistics, and, indeed, could be used to produce many other statistics of value that the government does not publish.

Senate Bill No. 1733, as passed in the chamber this week, provides that “the State recognizes the right of the people to information on matters of public concern and adopts a policy of full public disclosure of all of its transactions involving public interest.” It is plain to see that the government’s collection of data for statistical purposes is one such transaction involving public interest.

FOI includes raw survey data. Among the items included in SB 1733’s definition of “information” are the raw data vital for research: “…computer stored data, or any other like or similar data or material recorded, stored or archived in whatever form or format, which are made, received or kept under the control and custody of any government agency pursuant to law, executive order, rules and regulations, ordinance or in connection with the performance or transaction of official business of any government agency.”

Indeed, the National Statistics Office (NSO) does have a Data Archive of its raw data. But it lists only 64 “studies” from 1991 to 2010, which strikes me as rather few. Did the NSO do an average of only 3.2 studies per year over 20 years?  And where are all the pre-1991 works of the NSO—are they all lost?

FOI calls for the latest and most accurate data. SB 1733’s Section 9 on Openness and Transparency in Government Agencies states that each agency “shall regularly publish, print and disseminate at no cost to the public, and in an accessible form, consistent with the provisions of Republic Act No. 9485, or the Anti-Red Tape Act of 2007, and through their website, timely, true, accurate and updated information including, but not limited to, a list of ten items, including: “(8) Current and important database and statistics that it generates.”

As of now, however, the entire catalogue of the data archive of the NSO does not go later than activities of 2010. For instance, the public use file of the 2004 Updating of the List of Establishments was created on 6/18/2010,  but is still not available now, 10 years after the update, and almost four years after creation of the file. SWS is particularly interested in this update, for sampling establishments for its periodic survey of enterprises on corruption.

The public use file of the 2010 Census of Population and Housing was created on 6/26/2013, but is still not available, four years after the reference date and almost one year after creation of the file. For any survey organization, it is very important to have details of the latest census for sampling and weighting. Yet SWS, and any other institution without special access to the NSO, is forced to make do with the 2000 Census, which was done 14 years ago.

FOI should not be thwarted by overcharging. SB 1733’s Section 13 on Access and Processing Fees states that: “Government agencies may charge a reasonable fee to cover the actual cost of reproduction, copying or transcription and the communication of the information requested.”

Social Weather Stations had to pay the NSO P50,000 for a copy of the 2000 Census of Population and Housing, and P15,000 for a copy of the 2009 Family Income and Expenditure Survey. For one year of the Annual Poverty Indicators Survey, it appears that the charge is P8,000. Obviously, given the tiny cost of making a soft copy of electronic files, the NSO’s current pricing system is very exorbitant.

FOI for raw data can respect personal privacy. Section 7 of SB 1733 lists certain exceptions on disclosure, including: “(f) The information requested pertains to the personal information of a natural person other than the requesting party, and its disclosure would constitute an unwarranted invasion of his or her personal privacy. …” and “(g) The information requested pertains to trade secrets and commercial or financial information obtain from a natural or juridical person other than the requesting party, obtained in confidence or covered by privileged communication, and/or filed with government agency, whenever the revelation thereof would prejudice the interests of such natural or juridical person in trade, industrial, financial or commercial competition.”

These are part of the ethical responsibilities toward informants that survey researchers, whether in government or private practice, have always accepted. In particular: “No informant or respondent must be adversely affected as a result of his/her answers or of the interviewing process. (Rule 21, Code of Professional Ethics and Practices of the World Association for Public Opinion Research or WAPOR)”

Another rule is: “No response in a survey shall be linked in any way to an identifiable respondent. The anonymity of respondents shall be respected, except in rare cases, with the respondent’s specific permission.” (Rule 22, WAPOR code)

Survey archivists maintain these responsibilities by creating data files specifically for public use, meaning that they omit any information capable of identifying the respondent.

All in all, I think that SB 1733 is quite satisfactory, and I hope that the eventual version of the House of Representatives will be equally stringent.

* * *

Contact [email protected].