This report from the U.S. Office of the Inspector General (OIG), one of a number on State Department email and security processes conducted over the past five years, indicts archival practice more than it does the senior management of the State Department. Although it documents that secretaries of state since Madeleine Albright have ignored records management in going about their daily work, it shows that archivists, records managers, and cybersecurity experts issued impossible-to-follow guidance for electronic records and email for two decades and ignored evidence that their regulations were ineffectual.
The headline findings were that Secretary Hillary Clinton used a private server outside the department for unclassified email communications, which those policies would have permitted, without seeking approvals (though ironically, because of that, only Clinton's email has been saved). Of course, the report was big public news when released in May 2016, and the media widely reported and analyzed its conclusions with regard to Clinton,1 but overlooked the role of NARA because, like most archival failures, it wasn't very surprising. For archivists, the story wasn't so much that Hillary Clinton didn't follow rules in keeping her email, nor even that no one else in the State Department did either, it was that the rules proved useless and that NARA, the agency issuing the rules, whose job it was to preserve archival records, was hopelessly out of touch in its guidance and displayed feeble interest in enforcing it. It's not clear it even tried to figure out if it worked at all; according to the OIG, NARA “last reviewed the Office of the Secretary's record retention practices in 1991” (p. 14).
The bulk of the pages of this report attempt to identify the regulations that were supposed to guide federal agencies in their management of email and electronic security from 1997 to 2016. The State Department OIG does not question these regulations or directly challenge their implementability because that would have been outside its remit, but we, as a profession, should. Not only did these regulations obviously fail at State (and almost certainly everywhere else in the federal government), these failures led to the issuance of NARA's overly simplistic “Capstone” guidance in 2013 (which State did not adopt until 2015). The litany of these previous regulations reveals why archivists at the federal level have reduced their guidance to two superficial readings of the subtle function-based theory of macro-appraisal which are being understood as “save everything at the top of the agency” and “everyone else should save the things they think are important.” Now that the federal government has moved (in December 2016) from a default guidance of saving things by printing them out (which hardly anyone ever did [p. 17], and, had they, nearly all the important metadata would have been lost anyway), to saving things electronically (which few agencies are capable of doing successfully for more than a decade or so before format obsolescence and plain old system failures wipe out the records), everything proactively saved is likely to be at great risk of loss long before agencies get around to transferring the records to NARA. This is the state of archival guidance in 2017.
Sadly, in 1997 when this report first looks at the State Department Secretariat, archivists could have given useful advice and architected Business Acceptable Communications systems that would have captured all records at the point of their creation without depending upon the unreliable archival determinations of their authors.2 NARA rejected that approach, and the whole archival profession in the United States blindly followed a detour through diplomatics, attempting to apply medieval document verification methods to electronic communications and requiring potential records to be “set aside” after their creation to qualify for consideration as archives (as if enough humans could ever be found to assess the archival values of the many millions of e-documents created by employees of the federal government each day). Now NARA and the profession face the even greater embarrassment of having entirely abandoned the power of the principle of provenance as applied to appraisal by substituting a simplistic proxy of keeping all the electronic records of the higher-level officials of agencies (i.e., “Capstone”) and leaving the problem of their appropriate disposition to the future. Records scheduling is predicated on analysis of the functions of offices creating records, and records managers should have been analyzing these functions and implementing systems that automatically captured, and scheduled, records originating from these functions, in a recordkeeping system without relying on humans to set them aside or assess their significance. When American archivists lost sight of this basic premise, they lost any credibility they might have had to control archives in an electronic age. This is clear from the fact that in the dozens of news reports after the OIG report was released, I found none that sought the opinion of an archivist.
Meanwhile, other OIG reports in this series reveal that records of classified actions at the highest level of State are not being kept either.3 In all of 2013, only 41,749 email records were created in the State Messaging and Archive Retrieval ToolSet (SMART, the department's authorized recordkeeping system), out of more than one billion emails sent and received (pp. 3–4). Of these, 34,309 or over 82 percent, were from overseas missions, which in turn created them highly disproportionately and independently of their significance to U.S. policy: a consulate in Lagos made over 14 percent of the overall email records, while the embassy in Beijing made less than one record per week. State Department offices in Washington created a total of 7,440 records in a year, largely in administrative offices with information remits like the Bureau of Information Resource Management and the OIG itself, or about physical facilities (pp. 24–25). Critical policy-making offices are hardly represented at all: the Office of the Secretary under John Kerry created a total of seven records in SMART in all of 2013 (yes, a dozen or more professionals and the Secretary of State created about one record every fifty-two days), matching the number created by the Office of Population, Refugees and Migration (this in the year the Syrian refugee crisis exploded). This much bigger scandal has not become news, again because apparently no one cares. But archivists should.
Alas, it appears they don't care much either. As early as 2012, another OIG report concluded that State's “records management practices do not meet statutory and regulatory requirements. Although the office develops policy and issues guidance, it does not ensure proper implementation, monitor performance, or enforce compliance,”4 but no action was apparently taken prior to the subsequent reports documenting the consequences of this breakdown.
We could hope, now that the Clinton emails have raised the public visibility of these shortcomings, these reports will awaken the profession to its responsibilities, and archivists will find ways to insinuate themselves in records-making to ensure adequate recordkeeping in the federal government in the future, thus reclaiming some of their squandered public legitimacy. This will only happen once archivists acknowledge that all electronic documentation created in federal agencies should be treated from the moment of its creation as a record and therefore must be saved (if only for nanoseconds until its automatic disposal under an existing records schedule). To acknowledge this, American archivists must revisit a premise they rejected sometime around 1997 when this report first looks at emails of the secretaries of state and reopen a discussion of the functional requirements for recordkeeping that were active at that time.
Notes
1 For example, see Rosalind S. Helderman and Tom Hamburger, “State Dept. inspector general report sharply criticizes Clinton's email practices,” May 25, 2016, www.washingtonpost.com/politics/state-dept-inspector-general-report-sharply-criticizes-clintons-email-practices/2016/05/25/fc6f8ebc-2275-11e6-aa84-42391ba52c91_story.html.
2 David Bearman, “Item Level Control and Electronic Recordkeeping,” Archives & Museum Informatics 10 (1996): 195–245, http://www.archimuse.com/papers/nhprc/item-lvl.html; and David Bearman, “Towards a Reference Model for Business Acceptable Communications,” University of Pittsburgh School of Information Sciences (December 6, 1994), http://web.archive.org/web/19970707064048/http://www.lis.pitt.edu/~nhprc/prog6-5.html.
3 Office of Inspector General, Review of State Messaging and Archive Retrieval Toolset and Record Email (Report No. ISP-I-15-15, March 2015), https://oig.state.gov/system/files/isp-i-15-15.pdf.
4 Office of the Inspector General, Inspection of the Bureau of Administration, Global Information Service, Office of International Programs and Services (Report No. ISP-I-12-54, September 2012), 1, https://fas.org/sgp/othergov/ig-state.pdf.