MNCOGI background on records retention provided to Civil Law Committee February 2, 2017

OVERVIEW OF RECORDS MANAGEMENT

Prepared by Matt Ehling, MNCOGI board member

1.  The Records Management Statute (138.17) is separate from the Data Practices Act.  While the Data Practices Act regulates access to “government data,” the Records Management Statute (RMS) regulates how long government entities must keep “government records.”  Government records contain “government data” as defined by the Data Practices Act.

2.  The RMS does the following:

a.  Requires that government entities create “Records Retention Schedules” that specify how long they will keep government records;

b.  Empowers the Records Disposition Panel (the State Auditor, Legislative Auditor, Attorney General, and director of the Historical Society) to approve records retention schedules, approve when records can be destroyed, and transfer records with historical value to the State Archives at MHS;

c.  Defines various forms of “government records”;

d.  Regulates who has access to government records once they are transferred to MHS, since the Data Practices Act no longer applies once records are at MHS.

The RMS categorizes government records in terms of general categories of records, and not in terms of specific “form.”  For instance, the RMS speaks of “correspondence” and “memoranda” rather than “letters” or “e-mail.” It speaks of “recordings” instead of “videotapes” or “digital video files.”

3.  While the Data Practices Act does not apply to many rural townships, the RMS does.  However, the workload for rural townships is reduced, since they only need to maintain records under the RMS, but do not need to provide public access to the data under the Data Practices Act.

4.  The RMS requires government entities to create a records retention schedule, and to have it approved by the Records Disposition Panel.  Entities can either create their own, or use a model schedule, which many do.  MHS provides model records retention schedules that many cities, counties, townships, and other entities adopt.

The model schedules are available here.

For “correspondence” the model schedules specify the following retention periods:

a.  General County Retention Schedule:

General correspondence:  3 years

Medical examiner correspondence:  10 years

b.  General School District Retention Schedule:

Correspondence/administrative:  3 years

c.  General Township Retention Schedule:

General correspondence: 3 years

Historical correspondence:  Permanent

d.  General Cities Retention Schedule:

General correspondence: 3 years

Historical correspondence:  Permanent

Transitory correspondence:  Until read, unless content relates to city business and falls under other retention schedules.

5.  The Records Management Statute also interacts with the Official Records Act (15.17).  The Official Records Act states that government entities need to “make and preserve all records necessary to a full and accurate knowledge of their official activities.”  For many years, the Official Records Act was the only state statute that regulated government information.  John Finnegan and other people involved with the creation of the Data Practices Act disliked the Official Records Act because it did not provide much access to government information, and courts had interpreted it to be fairly narrow in its reach (see below).

6.  The Official Records Act was interpreted by a 1968 Minnesota Supreme Court case to only apply to “official” records.  The case (Kottschade v. Lundberg) held that government entities were not obligated to produce the field notes of a property assessor to the public, since such notes were not “official” records.  The enactment of the Data Practices Act – with its broad definition of “government data” – allowed the public to access government data such as field notes.  However, the concept of “official records” is still used by government agencies in the records retention area to categorize records that can be destroyed – even under a records retention schedule.  This is frequently applied to some types of e-mails. The thinking is that even if the public has access to a particular e-mail under the Data Practices Act, if that e-mail is not an “official record” then the entity does not need to keep it (unless it is the subject of a pending Data Practices Act request and disclosable as “public data”).  You can see that reflected in the recent St. Paul e-mail retention policy, where they state that messages that are not “official” can be destroyed outside of the records retention schedule.

MNCOGI background on records retention, presented to Sens. Limmer and Latz February 1, 2017

MINNESOTA COALITION ON GOVERNMENT INFORMATION

RECORDS RETENTION/E-MAIL – Background

Prepared by Matt Ehling, MNCOGI board member

The Records Management Statute (Minn. Stat. 138.17) governs retention of government records.

138.17 requires government entities to create a records retention schedule and to get it approved by the Records Disposition Panel.

The RMS categorizes government records in terms of general categories of records, and not in terms of specific “form.”  For instance, the RMS speaks of “correspondence” and “memoranda” rather than “letters” or “e-mail.” It speaks of “recordings” instead of “videotapes” or “digital video files.”

At present, there is no mandatory minimum time period for the retention of “correspondence” (including e-mail) under 138.17.

The Minnesota Historical Society (which serves as part of the Records Disposition Panel) provides recommendations for records retention.  Most of the recommendations for correspondence (including county, township, school boards) center on a 3-year recommended retention period for correspondence.  The model city schedule has a 3-year recommendation for “general correspondence” but allows “transitory” e-mails to be destroyed after reading.

The Records Management Statute interfaces with the “Official Records Act” at Minn. Stat. 15.17. The Official Records Act states that government entities need to “make and preserve all records necessary to a full and accurate knowledge of their official activities.”

15.17 has been interpreted to distinguish between “official” records and others that are not “official” and can be destroyed at will.  The Governor’s Office has been operating under such an interpretation since the Ventura administration, and more recently, the City of St. Paul adjusted its e-mail policy to limit record retention by highlighting the distinctions between “official” and non-official records.

MINNESOTA COALITION ON GOVERNMENT INFORMATION

RECORDS RETENTION/E-MAIL – Policy Recommendations

Eliminate references to “official” records in 138.17 and 15.17 and replace with “government records” to make the text of both statutes align, and to eliminate the artificial between “official” records and others.  This distinction currently allows government entities great discretion to destroy records, and should be eliminated.

To codify a minimum retention period for e-mail, set a minimum retention period for “correspondence” instead of “e-mail” in order to align with the existing language of 138.17.  If desired, “correspondence” can be further clarified to cover text-based electronic communications like e-mail.

Given that a 3-year retention period for correspondence is the most frequently recommended period in Minnesota Historical Society model schedules, MNCOGI urges the adoption of 3 years as the statutory retention period for correspondence, including e-mail.

To address concerns raised by government entities about retaining unnecessary amounts of non-pertinent e-mail (such as duplicate copies of e-mailed newsletters, etc.), provision can be made at 138.17 subd. (1)(b)(4) to specify certain specific categories of records that would not be subject to retention.

Pioneer Press: Any body camera law must be fully vetted

MNCOGI’s Don Gemberling is quoted in a Pioneer Press op-ed on body camera policy. The Pioneer Press editorial board notes that existing law should govern access to body camera footage until the legislature has enough time to study the nuances of the issue.

According to the editorial board:

“If lawmakers wait until more of them are better informed, as they should, existing law prevails. It stipulates that most footage body cameras collect is presumed public, except in certain situations — for example, while a criminal investigation is under way or if footage would reveal the identity of sexual assault victims.”

MNCOGI testimony: December 16, 2015 joint body camera hearing

MINNESOTA COALITION ON GOVERNMENT INFORMATION (MNCOGI)
Joint Hearing on body camera regulation, Dec. 16, 2015

Full written testimony of Matt Ehling
Chair, MNCOGI Legislative Issues Committee

Thank you for the opportunity to testify about Representative Scott’s proposed body camera bill.

The debate over how to regulate body cameras has been going on for over a year here at the legislature, and has raised many intertwined questions about personal privacy, government oversight, and technological novelty.

Over that year, people have asked whether body cameras are mechanisms for government transparency … whether they investigative tools … or whether they are invasions of personal privacy.

As technology continues to advance, these sorts of questions will get more and more involved, across all sorts of platforms. We feel that this bill is an example of a thoughtful approach that can be brought to these kinds of issues. It is a framework that still faces some adjustment, I’m sure, but it is a very solid conceptual base that addresses the many differing and competing aspects of the body camera debate.

Individual privacy
As I’ve mentioned, body cameras have both transparency and privacy impacts. I want to begin our comments by addressing how the bill deals with questions of personal privacy. This has been a recurring theme, due to the fact that body cameras are highly portable, and can easily enter private places such as homes.

Our organization has noted that by controlling when and how body camera video can be collected – especially in private places – the legislature can address some of the thorny privacy concerns that surround body camera use. And we’re pleased to see that this bill recognizes that approach.

Let me turn your attention to the second page of our packet. As you’ll see, in subdivision 3, the bill sets out a regulatory framework for the collection of body camera video. This framework deals with the use of body cameras in private places – places where there is a reasonable expectation of privacy.

Consent framework
And so to do that, the bill looks to the traditional framework that governs other police functions as a starting place for regulating body cameras. Fourth Amendment case law has long recognized that government actors must either have legal authorization to undertake certain actions, or they must have the consent of individuals. This bill takes that concept, and applies it to the use of body cameras. It recognizes that body camera use is separate and discreet from police entry onto property, and it requires its own, discreet treatment.

As you see in the packet, the bill requires that when police enter private property, they must first notify those they encounter that body cameras may be used. Then, they must also seek consent in order to record.

This consent-based approach applies to a wide range of situations that police face every day – welfare checks, responding to complaints, writing reports in private homes, and so forth. We believe that given the option, many people will choose not be recorded in private places during consensual encounters with police. If they do permit recording, their decision will have been made voluntarily, and with notice. And so the net effect of having a consent requirement in law will be to reduce the amount of video that’s recorded in private places, thus addressing privacy concerns, and also creating a smaller and more manageable pool of such data.

Exceptions
Now while consent will be the rule for body camera recording, there are several exceptions in the bill. Again, following the logic of Fourth Amendment case law, the bill recognizes situations in which police may record body camera footage without consent. Just like home entries and searches, consent to use body cameras is not required if police have a valid search warrant, or if exigent circumstances exist. The exigent circumstances category is well defined by case law that already governs entry and searches, and includes scenarios such as hot pursuit, intervening to stop violence, and so forth.

In addition, the consent requirement of this bill does not apply when police are responding to a report of domestic abuse, in recognition that the victim may be coerced, and may not be free to consent in those situations.

Data classification
Let’s now turn to questions of data classification – the other key part of this bill. Data classification has been an issue of dispute, and those disputes have centered around how much data to make public, or how much to make private. We have long contended that if police oversight is a key reason for why we have body cameras in the first place, then the resulting data needs to be classified to meet that purpose. This necessitates that – at minimum – data involving core police activities needs to be accessible to the public, and not just to the subjects involved.

Over the past year, we have advocated for fairly broad public access to body camera video. Others have disagreed, and have sought tighter controls. We feel that this bill finds a reasonable compromise that still preserves public access to the most critical oversight-related footage.

As you can see on page three of our packet, the bill breaks body camera data down into two distinct batches – data that document “emergencies, investigations, incidents, or request for services” – and those that do not. In our materials, we label the former activities – such as investigations – “core activities.”

Non-core activities
As you can see on page three, body camera footage that does not capture core activities is classified as “private” or “nonpublic” data – that is, data that is only available to the subject of the data, and not to the general public.

So what kind of footage might this be? One example would be footage from a body camera incidentally left on by an officer over their lunch break. Or from a body camera that was left running after a call for service had already been completed. That kind of footage would be classified as private data.

Core activities in public places
Moving down the page, let’s now turn to body camera footage related to core activities – those activities that have the highest oversight value. Again, this data involves “emergencies, investigations, incidents, and calls for service.” The bill classifies this data in two separate ways. First, if any of those activities occur in a public place, any body camera footage that captures them is classified as “public,” and is available through data requests.

Core activities in private places
Moving further down the page, body camera footage of core activities that’s recorded in private places – places where individuals have a reasonable expectation of privacy – is treated separately. That footage is classified as “private” data, with one key exception. The exception applies to footage that documents police actions that involve a use of force that results in bodily harm. And so this two pronged-classification makes most footage recorded in private places “private” data. But importantly – it also recognizes that even in private places, events may occur that have a public oversight value. For instance, in the event of a controversial police shooting, related body camera footage should be available to the public no matter if that shooting occurred inside or outside. The public interest in that footage is not diminished by the change of location.

Let me speak briefly to how this section interfaces with existing law. As with much other police data, data that’s made public by this section can be temporarily converted to a “not public” status during a criminal investigation. After the close of the investigation, the data would be public once again, but it would need to have image of protected persons – such as undercover officers – redacted before public release, as well as footage that is “clearly offensive to common sensibilities.” Both of those procedures are addressed by current law, and still interact with this proposed body camera section.

Other data-related provisions
There are some other data-related provision in this bill that I will briefly speak to as I close. The bill make a technical change to existing language in the Data Practices Act that I’ve already mentioned – language that allows material that is “clearly offensive to common sensibilities” to be redacted. Right now, that language says that “photographs” may be redacted. We’ve urged that the word “photographs” be replaced by “images” in order to cover body cameras and other recording devices. We are glad to see that the bill addresses this issue, and makes that change.

Finally, the bill makes it exceedingly clear that private vendors such as Taser International – the vendors that provide both the body camera systems and their data storage solutions – are covered by the Data Practices Act, including its provisions that allow for lawsuits and damages in the event of violations.

And so in closing, we are supportive of this proposal and what it seeks to achieve, and I’d be happy to take any questions.

Thank you.

Status of data in the Jamar Clark investigation

As has been widely reported in the local press, Twin Cities resident Jamar Clark was shot during an encounter with police in the early hours of November 15, 2015, and later died from that gunshot wound.  Controversy over the incident sparked demonstrations outside of the Minneapolis Police Department’s Fourth Precinct headquarters, and led to a protest-related shut-down of Interstate 94.

At present, both the Minnesota Bureau of Criminal Apprehension (BCA) and the Federal Bureau of Investigation (FBI) are conducting investigations into the shooting.  Much discussion in the press has focused on questions of when investigative data will be available to the public, in order to provide a clear picture of the incident.  MNCOGI presents the following summary of the status of such data:

Criminal investigative data:  When data gathered by police becomes part of an active criminal investigation, it becomes “confidential” or “protected nonpublic” data, depending on whether the data pertains to an individual or not.  While data is classified in this way, it cannot be shared with either the subject of the data, or with members of the public who ask for it.  This data classification extends to any video of the incident obtained by the police, with the exception of certain arrest or incident data.  After an investigation is closed, the “not public” data reverts to a public status, with certain exceptions.

Arrest data:  Even after a criminal investigation has been opened, certain data still remains public, even though other, related data gets converted into various forms of “not public” data.  Data that is public “at all times” includes certain arrest data, including data documenting whether any weapons were used by police, or whether there was any resistance encountered by police.  If this data exists in the form of a video recording, any portion of the recording that documents such details is public data.  In the Jamar Clark case, the BCA has indicated that it holds certain video of the incident.

Incident data:  Like arrest data, certain law enforcement “incident” data is always public, even after a criminal investigation has been opened.  Such data includes the date, time, and place of the police action, police report numbers, and other factual information (such as the use of weapons) similar to the “arrest data” discussed above.

Names of officers:  The names of the officers involved in the shooting were released on November 18th, several days after the incident occurred.  MNCOGI notes that the names of officers involved in any arrest or law enforcement incident are public “at all times” under Minnesota law.  This is true even if other data connected to an arrest or incident has been converted into criminal investigative data.

Public benefit data:  The law enforcement section of the Data Practices Act has a little-used provision that deals with “public benefit data.”  That section states that even when certain data is being maintained as “not public” criminal investigative data, police agencies can choose to release particular data to the public if the release would “aid with the law enforcement process, promote public safety, or dispel widespread rumor or unrest.”

Federal criminal investigative data:  Any Minneapolis Police Department (MPD) data related to the Clark case that has been transferred to the FBI will have its own classification under federal law.  Exemption 7(a) of the federal Freedom of Information Act prohibits the release of data on ongoing criminal investigations.  Federal law does not contain provisions that make some law enforcement data (such as certain arrest data) public during an investigation, as Minnesota law does.  To obtain arrest data related to the Clark case, it would be necessary to request it from the MPD or the BCA – not the FBI.

MNCOGI Supports Call For Legislative Openness and Transparency

Minnesota recently received a grade of D- from the Center for Public Integrity for its lack of transparency and openness in state government. In response, Minnesota House Minority Leader Paul Thissen has called for much needed and long overdue reforms. MNCOGI supports such reforms, and encourages Rep. Thissen and other legislative leaders to strengthen the public’s right to know.

The Center for Public Integrity and Rep. Thissen were both highly critical of a process where closed door meetings produced agreements and bills that were rushed to a vote before lawmakers knew what was in them.  Rep. Thissen said that the “chaotic conclusion” of the 2015 session was a “mockery of legislative process,” and called for reform legislation that would provide lawmakers and the public at least 24 hours to review any proposed legislation.

MNCOGI believes that this proposal is a good first step, but also believes that the legislature and the governor should do more.  Meetings where deals are being cut should be open to the public.  After all, it is the public’s business that is being done, and taxpayers’ money that is being spent.

Voters cannot hold their lawmakers accountable if they are kept in the dark, and the deal making is secret.  A little sunshine would go a long way to restoring Minnesota’s reputation for clean government.

If lawmakers cannot see their way clear to opening the doors, they should at least require that the closed meetings be recorded, and the recordings be made public after the legislation is adopted. That would provide the transparency the public needs in order to hold its elected officials accountable.

Back in the 1970s, the Minnesota Legislature passed good-government legislation called the Minnesota Government Data Practices Act.  The law applied to almost all levels of Minnesota Government.  Counties, cities, school districts, and more were included, but courts and the Minnesota Legislature were excluded. Rep Thissen thinks it’s time to see if the law should be extended to the legislature. MNCOGI believes that this is a sound suggestion, and pledges its support to craft appropriate legislation.

2015 legislative session round-up

James Shiffer (of the Star Tribune’s “Full Disclosure” column, as well as the MNCOGI board) uses his Sunday column to list the open government “wins” of the 2015 legislative session.  These include increased transparency measures for police surveillance equipment, as well as the rejection of bills that would have made certain body camera data, birth record data, and police incident data “not public.”  MNCOGI testified on all of these issues over the course of the session.  We will soon be posting our 2015 testimony and realted documentation in the “policy” section of our blog.  We use that portion of our site as a year-by-year archive of transparency issues that have arisen at the capitol.

MNCOGI notes that the legislative actions highlighed above were due to bi-partisan support from key, transparency-minded legislators, including Reps. Peggy Scott and John Lesch, as well as Senators Branden Petersen, Warren Limmer, Scott Dibble, and many others in both houses.  Our thanks go out to all who have supported the cause of government transparency in Minnesota.

MNCOGI opposes changes to birth record data

The Minnesota Senate is on the verge of approving a change to long-public birth record information.  A provision in the HHS omnibus bill would change address information on birth records to a “private” data status, from its currently “public” status under Minnesota law.  While the bill contains an exception for medical research, it would close off data to historical researchers, genealogists, and most other citizen requesters.

MNCOGI and the Minnesota Pro chapter of the Society for Professional Journalists recently co-signed a statement in opposition to this change – found below – and at the SPJ web site.

JOINT STATEMENT ON BIRTH RECORD DATA

The Minnesota Pro chapter of the Society of Professional Journalists and the Minnesota Coalition on Government Information (MNCOGI) oppose a provision of the Senate HHS Omnibus bill relating to the classification of birth record data. In particular, we oppose language (found in lines 217.24, 244.5-244.20, and 244.14) that would deprive the public of access to long-standing public information regarding birth records. For decades, such information has been useful to a wide range of Minnesotans for historical research, genealogy, and other lawful purposes.

By authorizing the release of birth record data only to limited sub-sets of individuals, such as “persons performing medical research,” these provisions of the omnibus bill advance the notion that data should be selectively released to the public based on the type of individual who requests it. This is at odds with the history of the Minnesota Data Practices Act, which does not discriminate between data requesters once government data has been classified as “public.” Such limitations could set a negative precedent for the subsequent classification of a wide variety of data in Minnesota. We urge the committee to reject proposed changes that would constrain public access to long-standing public information.

Sincerely,
Chris Newmarker (President, Minnesota SPJ)
Gary Hill (Chair, MNCOGI)

 

 

 

 

 

Open data and comedy: They can go together

MNCOGI board member Bill Bushey (center) shares the stage with Secretary of State Mark Ritchie (left) and Minneapolis Chief Information Officer Otto Doll (right)
MNCOGI board member Bill Bushey (center) shares the stage with Secretary of State Mark Ritchie (left) and Minneapolis Chief Information Officer Otto Doll (right)

 

Open Twin Cities co-founder and MNCOGI board member Bill Bushey braved the stage at Bryant Lake Bowl on Monday to sing the praises of open data, along with the Minnesota Secretary of State and Minneapolis’s chief information officer. It was all part of the Theater of Public Policy, an improv group that combines comedy with free-wheeling discussions of serious stuff. The house was packed to hear about the availability of large data sets, privacy breaches and other typically dry matter made more fluid by the wide beer selection and live music. Also spotted in the audience: MNCOGI board member Helen Burke.

City finally hands over records of investigation

By James Eli Shiffer, MNCOGI board member and the Star Tribune’s watchdog and data editor

I blogged last month about how the Star Tribune had waited seven months for the city of Minneapolis to hand over records of an internal investigation into a public official. The documents finally arrived last week, and staff writer Eric Roper had a story about it in Sunday’s paper.  He also offered readers the entire investigative file, so you can piece together your own story, in between eight months’ worth of redactions.