This article is Part 4 in Linda Pannozzo’s series: Biomass, Freedom of Information and the Silence of the DNR Company Men. The proceeding articles are:
Part 1: Reporter Linda Pannozzo discovers just how hard provincial bureaucrats worked to ignore her questions.
Part 2: An Open Letter to the FOIPOP Review Officer
Part 3: What Happened When This Reporter Got Called Down to the Office
“I feel like I’ve just been given the total run-around,” is how I began my final correspondence with Mary Kennedy, the Intake Manager at the Office of the Information and Privacy Commissioner (OIPC) for Nova Scotia.
About a year ago I contacted the Department of Natural Resources looking for information about the forests in this province — specifically about the age classes and species composition. When nothing was forthcoming through its media relations person Bruce Nunn, I decided to file a Freedom of Information (FOI) request, which was very specific. I wanted “raw data for the most recent GIS forest inventory for age class and species composition of the Province and regions.” 1 I figured that while I was at it I’d also request the age class and species composition data derived from the Permanent Sample Plot measurements, which are highly suspect to say the least — and a subject I’ll return to.
The initial FOI request was answered by then DNR Deputy Minister Frank Dunn. He wrote that the GIS-derived age class data were “a matter of public record” and available to the public by entering a “Data Sharing Agreement.”
But as it turns out, what I got back from the DNR was false information, something I wasn’t aware of at the time that I wrote the Open Letter to the FOIPOP review officer, Catherine Tully requesting a review of Dunn’s letter.
Now, in case you haven’t been following this series, let me back-track a bit to bring you up to speed.
The reason I asked for a review was because I was really frustrated. Since the 1950s the DNR has reported on age class and species composition for the province and regions in periodic “forest inventory” reports, where reams and reams of data were made decipherable as information for public consumption. The historical trend over a 45-year period was showing that the province’s forests were getting much younger, old forests were disappearing, and that species composition was changing — turning what used to be Acadian forest into the boreal. I wanted to update the historical trend line for age class (see chart below) with the most recent data but according to Dunn, the only way to access it was by signing a Data Sharing Agreement.
But having to sign an “exclusive” agreement with a government body over what should be public information is fraught with problems, all of which I outlined in the open letter.
But something else was a little fishy about Dunn’s letter. It said the GIS-derived age class data — the source of the previous 45 years of historical information — was included in the Data Sharing Agreement, but when I got a sample copy of the agreement I noticed that it wasn’t. Only PSP-derived age data was.
So I contacted the DNR and asked them about that and in an email, Forestry Division head Jonathan Kierstead (formerly with Bowater Mersey) pointed me instead to the government’s newly launched “open data portal.”
While Tully’s office was reviewing my file, the DNR invited me down to its office on Hollis Street in Halifax. They wanted to “explain the data” and “processes” and address the “challenges” I’d been having in accessing information. So, in July I met with Allan Eddy, the former Associate Deputy Minister of the DNR, Chris Bailey, who works with forest inventory, and Dan Davis, the head of Communications. 2
I wrote about my summer meeting with the DNR in Part 3 of this series, and was informed — as Kierstead had already advised — that if I wanted the GIS age class inventory data I could get it through the open data portal. To my dismay, however, the government’s idea of transparency would require that I install a bunch of software just to open the file extensions and then hire a forestry consultant to interpret the numbers for me.
Now I’ll be honest, I dug in my heels… until Kennedy’s next letter.
In September she wrote saying that the kind of accessible, report-style inventory information I was accustomed to (and which the government had been providing the public for nearly half a century) no longer existed:
During the course of this review, in response to questions regarding the requested information, DNR advised the OIPC…The requested information is only available electronically through Open Data and through Digital Geographic Data via the Data Sharing Agreement. It can be accessed by downloading the shape files (opened by downloading ArcExplorer) and by navigating the Provincial Landscape Viewer to visually see the Forest Inventory presented in layers… there are no fees associated with accessing this data.
Kennedy said that the Freedom of Information Act did not apply to the records I requested because, a) the information was a matter of public record, and b) the information is published — both of which are excluded from the Act. She also made it clear that even if the Act were to apply to requested information, it did not require a public body to present information in a format that was more understandable to the public. Kennedy was about to close my file, but I had 15 days to respond if I wanted it to remain in the queue to be investigated by the Commissioner herself, though she warned that the Commissioner would likely “come to the same finding.”
I decided to follow the process Kennedy outlined and I downloaded the software and then went to the open data portal and downloaded the “forest inventory” data, but was still unable to open any of the file extensions (.dbf, .prj, .sbn, .shx. and .shp). I saved the data on an external hard drive and decided that my next step would be to bite the bullet and hire someone who could, a) open it, and b) make sense of it for me.
But then I noticed something strange: age class wasn’t listed as one of the “forest attributes” on the open data site.
Before my 15 days were up I contacted Kennedy and told her what had happened — that the GIS-derived age class data were not where the DNR said they were. She then contacted the DNR and this is part of what she relayed to me:
Age class is not an attribute of the inventory files… records of age class do not exist as DNR does not use age class to define forests… it would appear that this information does not exist.
So this brings us full circle. In my last email correspondence with Kennedy, I wrote:
I feel like I’ve been misled from the very beginning. The DNR would have known back in April that GIS-derived age class data were not available through the Data Sharing Agreement. But that was what the letter from [Frank] Dunn said. They also knew age class attributes were not available through the open data portal, and yet, [on two occasions] they told me they were. What recourse do I have at this point?
Apparently none. Kennedy closed my file in December with a letter stating:
The only recourse to address this issue would be to recommend to the public body [the DNR] that it be clearer when responding to applicants about whether or not a requested record actually exists, which recommendation I have already made to DNR. There is no further remedy available through this process.
So I guess the review is closed, but the case isn’t.
Back in 1954 the journalist Darrell Huff wrote the best-seller How to Lie with Statistics — a book meant to introduce readers to the world of statistics and how they can be used erroneously to mislead or distort reality. In short, how numbers can be used to deceive you.
Huff presented example after example of how statistical material was used to disinform and manipulate, with the intent to deceive. He provides the reader with five questions they should always ask when presented with any statistic: Who says so? How does he know? What’s missing? Did somebody change the subject? Does it make sense?
While the kind of shenanigans the DNR seem to be engaged in here isn’t dealt with specifically in Huff’s book, it seems to fall under the general category of what he calls “statisticulation” or statistical manipulation.
And the way the swindle seems to be playing out in this case is relatively concealed — by changing definitions and creating breaks in data series. For instance, when you use the same term but change what it means, it can result in a sudden (but false) improvement in what’s being tracked. The clearcutting definition is a prime example of that.
In 2010, the NDP government committed to a target of reducing clearcutting to 50 per cent of all harvesting, which received all party support as well as public support. Then in August of 2012 the Dexter government released a new definition of clearcutting: “where less than 60 per cent of the area is sufficiently occupied with trees taller than 1.3 m.”
Forester and now lawyer Jamie Simpson said at the time that in practical terms, this new definition meant that “half of all cutting can leave a moonscape; the other half can leave a scattering of low-quality trees, none necessarily higher than four feet.”
Changing the clearcutting definition would make it a lot easier to reach the target without actually having to change harvesting practices.
But the new definition didn’t get much traction. First, it was hard to explain, particularly when the National Forestry Database (mandated through the Canadian Council of Forest Ministers) continued to report that clearcutting in Nova Scotia was in the vicinity of 90 per cent. And, as I reported here, the new definition wasn’t really needed anymore when the DNR reported in its progress report on the fifth anniversary of the 2011 Natural Resource Strategy that it had abandoned a whole slate of citizen-led targets including the 50 per cent reduction.
So when it comes to age class we need to ask several questions:
• If the government no longer uses it to define forests, then what is it using and where is the underlying data coming from?
• What is the rationale behind changing how it defines forests?
• And how would the change affect the information we’d be getting about the state of the forests, if we were to actually get any?
First off, when you stop using age class to define a forest it means you can no longer update historical trend lines because the numbers of the past — derived differently and meaning different things — cannot be compared to the current numbers, creating a “break” in the data series. In other words, long term trend lines — precisely what are required if claiming sustainable forest management — are no longer possible. The forests of today can no longer be compared with the forests of the past.
The other issue that should be addressed is: Which data are being used to describe the forests? It appears, from the little information I’ve been able to obtain from the DNR that data from Permanent Sample Plots is taking up the slack. This from DNR media relations, Bruce Nunn:
The GIS inventory data is used for landscape level planning and does not have the precision of the PSP data. PSP data is based upon field measurements whereas the forest inventory GIS data is based upon photo interpretation…Precisely quantifying age and development class over time using the forest inventory GIS data would not be the appropriate use of this data set. The PSP data is the appropriate data set to answer questions regarding the change in forests conditions over time. This is a common practice across the country.
But what do we know about PSP data? We know that the DNR maintains more than 3,000 of these randomly placed plots across the province. We know that the system has been in place since 1965, but according to the DNR it was always “designed to track volume growth and mortality of the natural forest.” The first time it was ever used as a forest inventory was 1994-1998, so it lacks historical depth. But the other, more crucial problem with these plots is no one seems to know if they are in fact representative of anything.
In order to extrapolate the measurements on these plots to the province as a whole, it must be assumed that the plots are treated in exactly the same manner as the surrounding lands and not given preferential treatment. In other words, if the land is clearcut around a PSP plot, then the PSP plot has to be clearcut too. But the numbers — as well as off-the-record interviews with people in the field — seem to suggest this is not the case. PSP data show higher volumes and proportions of old trees, which indicates that the plots might not be harvested at the same rate as the adjoining lands, resulting in exaggerated numbers. For example, when GIS data was indicating that less than one per cent of the forests were between 80 and 100 years, the PSP plot data was saying it was closer to 14 per cent. That’s a significant difference.
The way to verify that PSP data are in fact representative of NS as a whole would be to conduct a field study/ audit (independent or internal) on the plots. But according to the DNR there has never been an audit on the permanent sample plots.
What is most troubling about this is that according to the DNR, the PSP data is what is currently being used to assess changes in forest condition over time and to “improve the volume estimates within wood supply models.” If the PSP data are as exaggerated as I believe them to be, then this could also result in an exaggerated wood supply.
According to the DNR, “age classes” have now been replaced with what they call “development classes” and “seral stages,” because they “best represent the current state of the forest and how we want to manage it today and into the future.”
In an email received just before this article went live, DNR stated that the seral stages and development classes were a “better tool” than using “age class.” They sent me this 2008 document, A Procedural Guide for Ecological Landscape Analysis: An Ecosystem Based Approach to Landscape Level Planning in Nova Scotia, which provides the definitions for the two terms. 3
But here’s the point. There is nothing wrong with using these terms — indeed they are concepts used in landscape ecology where forests are treated as inherently valuable ecosystems, a perspective that is crucial to our survival on this planet — but using them also implies that some semblance of ecologically-based forest management is being employed. The problem is, it isn’t. When it comes down to how the forests are being degraded in this province — by the pulp and paper and now biomass industries — there’s nothing ecological about it.
In an interview conducted before he left his Forestry Coordinator position at the Ecology Action Centre to take on a position at WoodsCamp, Matt Miller said the DNR don’t actually want anyone to know the real age class distribution “because it’s damning.”
Miller explained how the government’s new ecological landscape approach led to changing other definitions as well. One of the new “development classes” is the “mature” class and it’s now defined as anything older than 40 years old, while the EAC defines it as anything over 80, said Miller. Changing that definition allows the DNR to include more (younger) forests in the “mature” category, which also sounds good on paper.
For nearly 50 years the DNR showed us the numbers despite the fact that they pointed to some very worrying trends. In the late 1990s the government warned that we were liquidating our forests and harvesting on private lands was unsustainable. Since then, despite overwhelming public sentiment for a sea change in forest management the government has instead been busy with something else — working on behalf of industrial forest interests — all under the guise of ecology and forest science.
Linda Pannozzo is a freelance journalist and author. Her latest book is About Canada: The Environment, published in November 2016 by Fernwood.
- GIS stands for Geographic Information System and it’s a type of forest inventory data that the DNR has been collecting since 1958. The DNR has been collecting data on the “age classes” and “species composition” of forest in this province since 1953. Historically, data were first collected using field strip cruises and these were replaced with field sampling and photo interpretation and the more recent inventories, since 1985 were derived from aerial photographs and photo interpretation of roughly 2 million forest stands, which were then verified by field sampling. ↩
- It should be noted here that in early November 2016, The Examiner reported that Allan Eddy was involved in muzzling Robert Cameron, a Department of Environment lichen scientist and by the end of the month, in a move that may or may not have been connected, the government announced that both Eddy and Dunn would be moved from the DNR to Agriculture and Fisheries and Aquaculture, and Julie Towers—the partner of the muzzled lichen scientist—was promoted from her previous position as CEO of Aboriginal Affairs and Immigration to Deputy Minister of DNR. ↩
- From the 2008 document: “development classes describe changes in structure and process as forests age” and that for landscape management purposes, four development classes have been recognized: forest establishment, young forest, mature forest, and multi-age/ old growth forest. Seral stages, according to the document, describe the successional changes that take place in a forest over time and that three seral stages are recognized: early, mid, and late. ↩