Greenpeace’s Fun with Graphs

by on December 4, 2007 · 24 comments

Greenpeace has released the latest edition of its quarterly Guide to Greener Electronics. While I haven’t read the study in full and I don’t know exactly what goes in to determining the one through ten ranking that Greenpeace assigns to various famous tech companies, I did find their graph (see below) a little odd. Look how close together one to three are! Then look at the space between seven and ten–it’s half the graph! By making three numbers take up half the graph, a greening tech company can move quite a way across the “dial o’ green” if it moves from a seven to an eight, but a move from three to our doesn’t result in such a pronounced leap.

Adopting cleaner technology standards and practices is important, don’t get me wrong. But such a blatantly misleading graph makes me question the legitimacy of this entire quarterly report. Can we get some unbiased research into this area of tech please?

NOTE: Some comments have shown me that I wasn’t clear in the original post just how manipulative these graphs are with the data. It’s important to note that past graphs show a rank of 5 as the midpoint of the graph. The most recent graph shows a rank of 7 as the midpoint. This way, companies that have actually gotten greener appear to be back-sliding.

  • http://www.techliberation.com/ Tim Lee

    Isn’t the scale arbitrary anyway? I don’t think it’s misleading to graphically re-scale an arbitrary index. There’s no particular significance to the number 5.

  • http://www.techliberation.com/ Tim Lee

    Isn’t the scale arbitrary anyway? I don’t think it’s misleading to graphically re-scale an arbitrary index. There’s no particular significance to the number 5.

  • PJ Doland

    I believe the headline you were looking for was “This One Goes to Eleven.”

  • PJ Doland

    I believe the headline you were looking for was “This One Goes to Eleven.”

  • http://www.cato.org/people/harper.html Jim Harper

    It looks like they move the goalposts from version 1-3 to versions 4-6. A consistent graph would appear to show improvements. Changing the scale of the graph suggests to the careless viewer that there hasn’t been progress and leaves most of these companies lumped “in the middle.”

  • http://www.cato.org/people/harper.html Jim Harper

    It looks like they move the goalposts from version 1-3 to versions 4-6. A consistent graph would appear to show improvements. Changing the scale of the graph suggests to the careless viewer that there hasn’t been progress and leaves most of these companies lumped “in the middle.”

  • http://www.cato.org/people/harper.html Jim Harper

    And I concur with PJ.

  • http://www.cato.org/people/harper.html Jim Harper

    And I concur with PJ.

  • http://www.cato.org/people/harper.html Jim Harper

    Note that while the numerical scale moves to the left, the color scale stays the same. So Dell, a well-green 7 in August 2006, has moved to the left and fallen back to reddish-green in December 2007, even though it has improved to 7.5.

    And I concur with PJ.

  • http://www.cato.org/people/harper.html Jim Harper

    Note that while the numerical scale moves to the left, the color scale stays the same. So Dell, a well-green 7 in August 2006, has moved to the left and fallen back to reddish-green in December 2007, even though it has improved to 7.5.

    And I concur with PJ.

  • http://www.cordblomquist.com Cord Blomquist

    Check out HP. It has made significant strides and has gone from a 5 to a 7, but remains in the same place on the graph.

    Tim, you have a good point. If Greenpeace’s goal is make these companies look like they’re under-performing, they should just arrange the scale in a different way. Though I do enjoy the fact that instead they just moved the numbers around on the graph, it’s the more telling way that Greenpeace is trying to manipulate the results.

  • http://www.cordblomquist.com Cord Blomquist

    Check out HP. It has made significant strides and has gone from a 5 to a 7, but remains in the same place on the graph.

    Tim, you have a good point. If Greenpeace’s goal is make these companies look like they’re under-performing, they should just arrange the scale in a different way. Though I do enjoy the fact that instead they just moved the numbers around on the graph, it’s the more telling way that Greenpeace is trying to manipulate the results.

  • http://www.techliberation.com/ Tim Lee

    Oh, you’re right, they’re moving the goal posts from version to version. I don’t see a problem with numbering the graph in a non-linear fashion, but it is sleazy to move the numbers around to mask the fact that companies are improving.

  • http://www.techliberation.com/ Tim Lee

    Oh, you’re right, they’re moving the goal posts from version to version. I don’t see a problem with numbering the graph in a non-linear fashion, but it is sleazy to move the numbers around to mask the fact that companies are improving.

  • http://enigmafoundry.wordpress.com eee_eff

    I find nothing inappropriate or misleading about the graph–using a log or geometric scale is a frequent tool is graphs (graph paper companies even make log and geometric grids*) in Scientific literature, for example.

    Such a graph is actually better here because all the criteria are NOT all equal–some are relatively easy to achieve while others are much more difficult, and it makes very good sense that most companies will try to do the ‘low hanging fruit’ first, so a linear graph wouldn’t make sense.

    I would also note that Greenpeace is very transparent about their criteria, which can be found on their website, and their methodology, in as much each particular rating given any company is explained, in detail, criterion by criterion.

    That being said, the rating system is certainly far from perfect–its open to question whether other things those companies are doing might off-set their other unsustainable practices.

    So, the “blatently misleading” charge that Cord makes is, in fact way off base.

    The observation I would have is that today, in the present state of the environment, we need to raise our standards and expectations, not lower them, and I applaud Greenpeace for doing this.

    *At least they did when I took physics in college–perhaps with the advent of computers, log scale graph paper has gone the way of the slide rule.

  • http://enigmafoundry.wordpress.com/ enigma_foundry

    I find nothing inappropriate or misleading about the graph–using a log or geometric scale is a frequent tool is graphs (graph paper companies even make log and geometric grids*) in Scientific literature, for example.

    Such a graph is actually better here because all the criteria are NOT all equal–some are relatively easy to achieve while others are much more difficult, and it makes very good sense that most companies will try to do the ‘low hanging fruit’ first, so a linear graph wouldn’t make sense.

    I would also note that Greenpeace is very transparent about their criteria, which can be found on their website, and their methodology, in as much each particular rating given any company is explained, in detail, criterion by criterion.

    That being said, the rating system is certainly far from perfect–its open to question whether other things those companies are doing might off-set their other unsustainable practices.

    So, the “blatently misleading” charge that Cord makes is, in fact way off base.

    The observation I would have is that today, in the present state of the environment, we need to raise our standards and expectations, not lower them, and I applaud Greenpeace for doing this.

    *At least they did when I took physics in college–perhaps with the advent of computers, log scale graph paper has gone the way of the slide rule.

  • Charles

    EF,
    I think what you’re missing (and I’m not blaming you, I didn’t realize this at first either and I don’t think Tim did either) is that when you flip through the different “versions” of the graph, the axis get rescaled, but not equally.

    Flip to the version 3. Zero on the extreme left and 10 on the far right, with all the spacings being equal. Now flip to version 4 and, magic, the endpoints are still the same, but, the spacings are different. 1, 2, 3, 4, 5 are all small and 5 6 7 8 9 10 get progressively bigger. So companies that showed an improvement actually move to the left on their plot.

    That’s misleading.

  • Charles

    EF,
    I think what you’re missing (and I’m not blaming you, I didn’t realize this at first either and I don’t think Tim did either) is that when you flip through the different “versions” of the graph, the axis get rescaled, but not equally.

    Flip to the version 3. Zero on the extreme left and 10 on the far right, with all the spacings being equal. Now flip to version 4 and, magic, the endpoints are still the same, but, the spacings are different. 1, 2, 3, 4, 5 are all small and 5 6 7 8 9 10 get progressively bigger. So companies that showed an improvement actually move to the left on their plot.

    That’s misleading.

  • http://enigmafoundry.wordpress.com eee_eff

    Charles:

    The way the animation works on the Greenpeace home page, that is: starting from zero and dynamically reposting each companies location on the chart makes comparisons between years difficult, so it’s very hard to compare where say, Apple is one year to where Apple was the next year, because that’s not the intent. I think the case that these graphs are misleading is way overstated, although from a purely technical point of view you have found an inconsistency between editions, and they should have implemented the archive function differently–so each appeared in a new window, perhaps, so it wouldn’t invite the comparison you are trying to make.

    The graphs, by the way they are animated, invite comparisons between different companies in any given reporting period, which is the graphs were intended to work: so you can compare different companies.

    Moving the goal posts is somehow being portrayed as unfair, but it is necessary to raise our expectations on all issues related to sustainability. Changing the metrics by which we measure sustainability is going to be a permanent, major feature and driver of the necessary transformation of all industries, in all countries, for the foreseeable future. The economy has to adapt to the very real limitations of the environment, and that process will be highly dynamic.

    If you think these goal posts have moved you ain’t seen nothing yet–for example the American Institute of Architects has adopted the 2030 challenge which calls for

    “- a 50 percent reduction in fossil fuel energy consumption—over that of an average building of the same type in the same area— in new buildings by 2010

    - 0 percent carbon emissions in new buildings—by 2030.”

    That’s right folks, all new buildings should be at least carbon neutral by 2030. How’s that for moving goalposts? My advice: get used to it, i.e., adapt or die.

    Rating systems have been shown to have dramatic effects in driving market changes towards social goals, witness the effect that the United States Green Building Council’s LEED rating system has had on the market for building materials and building design/engineering services. It’s changed that industry 180 degrees…

    The Green Lever of Trademarks

  • http://enigmafoundry.wordpress.com/ enigma_foundry

    Charles:

    The way the animation works on the Greenpeace home page, that is: starting from zero and dynamically reposting each companies location on the chart makes comparisons between years difficult, so it’s very hard to compare where say, Apple is one year to where Apple was the next year, because that’s not the intent. I think the case that these graphs are misleading is way overstated, although from a purely technical point of view you have found an inconsistency between editions, and they should have implemented the archive function differently–so each appeared in a new window, perhaps, so it wouldn’t invite the comparison you are trying to make.

    The graphs, by the way they are animated, invite comparisons between different companies in any given reporting period, which is the graphs were intended to work: so you can compare different companies.

    Moving the goal posts is somehow being portrayed as unfair, but it is necessary to raise our expectations on all issues related to sustainability. Changing the metrics by which we measure sustainability is going to be a permanent, major feature and driver of the necessary transformation of all industries, in all countries, for the foreseeable future. The economy has to adapt to the very real limitations of the environment, and that process will be highly dynamic.

    If you think these goal posts have moved you ain’t seen nothing yet–for example the American Institute of Architects has adopted the 2030 challenge which calls for

    “- a 50 percent reduction in fossil fuel energy consumption—over that of an average building of the same type in the same area— in new buildings by 2010

    - 0 percent carbon emissions in new buildings—by 2030.”

    That’s right folks, all new buildings should be at least carbon neutral by 2030. How’s that for moving goalposts? My advice: get used to it, i.e., adapt or die.

    Rating systems have been shown to have dramatic effects in driving market changes towards social goals, witness the effect that the United States Green Building Council’s LEED rating system has had on the market for building materials and building design/engineering services. It’s changed that industry 180 degrees…

    The Green Lever of Trademarks

  • http://enigmafoundry.wordpress.com eee_eff

    Oh, I think I figured it out. The graph also jumps a little from edition 4 & 5 to edition 6.

    What appears to be happening is that the mid point of the companies scores stays in the exact middle of the graph.

    So Greenpeace isn’t moving the graph the companies themselves are.

    Greenpeace is just doing what my high school math teacher did: grading on a curve!

    The complaints from the under-performing students didn’t cut it back then, and the complaints of the underperforming companies don’t cut it now either.

  • http://enigmafoundry.wordpress.com/ enigma_foundry

    Oh, I think I figured it out. The graph also jumps a little from edition 4 & 5 to edition 6.

    What appears to be happening is that the mid point of the companies scores stays in the exact middle of the graph.

    So Greenpeace isn’t moving the graph the companies themselves are.

    Greenpeace is just doing what my high school math teacher did: grading on a curve!

    The complaints from the under-performing students didn’t cut it back then, and the complaints of the underperforming companies don’t cut it now either.

  • http://pj.doland.org/ PJ Doland

    After looking at this again, they might have just done this so they could expand the area between 7-8 to fit all the company names in there.

    Version 1 only has 14 companies but version 6 has 18.

  • http://pj.doland.org/ PJ Doland

    After looking at this again, they might have just done this so they could expand the area between 7-8 to fit all the company names in there.

    Version 1 only has 14 companies but version 6 has 18.

Previous post:

Next post: