“Half of single-use plastic waste is produced by just 20 companies."
So proclaims the headline for a 5/18/21 CNN.com article. This isn't the only place you could read about it. CNN, NPR, BBC....they all bought the same report from Reuters, used roughly the same headline for the same story.
What does the statistic in this headline tell us? Is it accurate? Why should we care?
To be clear, I'm not asking whether we should care about single-use plastic waste. We should. Roughly half the world's plastic waste is single-use, and it's increasingly poisoning the earth and its inhabitants (more details here).
What I'm asking is why we should care that X percent of single-use plastic waste is produced by Y companies. What's at stake here? When does it matter what X and Y are? Is this just a dry, academic question? (Hint: I don't think so.)
Questions like this are the focus of my newsletter. Welcome to the first installment!
Statistics as we know it is a recent invention. (In a future newsletter I'll unpack that phrase "as we know it.") It deserves part of the credit for unprecedented advances in science, technology, health care, education, social policy, etc.
At the same time, statistical information is easily misunderstood and misused. Partly because it's technical. And partly because John Q. Public isn't the only one who gets it wrong. Journalists, politicians, doctors, sportscasters, and bloggers misrepresent statistical data. Governments and policymakers do it too. Scientists themselves misuse the very statistics on which their research is grounded. What these folks have in common is that regardless of profession, the misuse of statistics is sometimes unintentional, sometimes malicious.
One of the ongoing themes of this newsletter will be that statistics can guide us and lead us astray. I hope this newsletter will be a good compass.
(If you know something about statistics, you may be wondering: When I use the word "statistics", do I mean the academic discipline, the mathematical assumptions and procedures we call statistics, or the actual numbers generated by these procedures? Well, all of them. I'll try to be clear about which I'm referring to whenever I use the term.)
Each week I'll be taking a close look at some data I've encountered that week – typically in the news, in social media, in advertising, or in a conversation. I'll tell you where the data came from (often, a new study), I'll discuss its credibility, I'll share my thoughts on what to do with it, and I'll add some remarks about the broader role of statistics in contemporary society. Health, parenting, education, psychology, politics, and the environment are among my interests, but I expect to cover other topics as well.
So... let's return to that headline.
Even if you know nothing about single-use plastic production, as an educated, 20th century American you get the headline's implicit message: A small number of companies are producing a large percentage of the waste. Crudely speaking, the headline is playing an unfairness angle to get you to read the story. We see headlines like this all the time. ("The top 1% of Americans now control 38% of the wealth" cnbc.com.)
Of course, CNN could've just said: A minority of companies produce the majority of the waste. But that wouldn't work as a headline, because it would sound vague to our 21st-century ears. This brings me to my first point about the headline: It illustrates how statistics serve a legitimizing function. We cite statistics to make things sound more legitimate, more credible, more accurate. This isn't a bad thing, but it's new, historically speaking, and it has some downsides.
In this particular instance, it's hard to tell how to read the statistic without further information. The CNN article throws around a lot of numbers, but here's one that's missing: How many companies produce single-use plastic waste? You can't find that number in the NPR and BBC articles either. If the answer is 30, then saying that half the waste is produced by 20 of these companies isn't particularly noteworthy. If the answer is 30,000, then 20 would be startling indeed. In fact, the report from which this statistic was drawn refers to approximately 300 polymer producers worldwide. 20 of these 300 companies produce slightly more than half of all single-use plastic waste, according to the report. 100 of these companies produce 90 percent of all single-use plastic waste.
Now that we have some context, the question still remains: What should we do with this statistic? I would argue: Not much. The report offers advice on how to address the broader problem. Like pressing for transparency about who produces single-use plastics. (That grocery store bag probably doesn't have the producer's name printed on it.) And holding companies as well as the financial institutions that support them accountable, putting pressure on them along with policymakers to shift toward production of recyclable plastics. (Of course, there are problems with recyclables, one being that in some cases they're not truly recyclable – check out John Oliver's 3/21/21 monologue on plastics.) In short, what we should do to reduce the environmental impact of plastics has almost nothing to do with how many companies produce what percentage of single-use waste.
I say "almost" because one could imagine a different scenario. For example, if the #1 producer of single-use plastic waste – ExxonMobil – was producing 97% of it, then advice for dealing with the problem would likely be different (i.e., highly focused on one company). However, it turns out that ExxonMobil only produces 5.9% of the total waste, meaning that individual, company-targeted strategies (like consumer boycotts) might not be as efficient as, for example, calls for new legislation around recyclable plastic production. Again, whether it's 20, or 30, or 40 of the 300 companies producing more than half the waste doesn't seem particularly important.
So....why even bother picking on this statistic? Because it illustrates that our reliance on statistics for legitimacy has become so deeply ingrained, we trot them out even when they're not meaningful or grounded in evidence. ("We" meaning not just the guy at the bar or the sportscaster, but even national news agencies.) Here's one of my favorite examples: On the back of my "purely elizabeth" brand bag of granola (made of recyclable plastic) I find the following piece of advice:
"Follow the 80/20 rule. 80% of the time be your healthiest self. 20% of the time, indulge – guilt-free."
In some future newsletter, I will discuss the problems with statistics like this. (You've probably thought of a few already.) Here, I just want to call attention to the framing of this lifestyle advice in statistical terms, as if the stats somehow made the advice "better" – more credible, more scientific, or whatever. I don't expect I'll be changing my life based on this "statistic". But I am interested in the assumption that I might've done so, just because numbers were provided. As for that 20-out-of-300 polymer producers figure...it's important information (mostly for telling us that it's not 1 or 2 companies who are outliers), but what I care about more is doing what I can as a citizen to reduce single-use plastic waste. (For further information on what you can do, check out the report here.)
Thanks for reading!
Mine is not as long as one of my favorite people, Sumei but I have long thought of the way statistics are thrown around. Many times I have thought they were used for benefit of those using them rather than to give a truthful view of the topic being discussed.
Wow, how incredibly insightful it is to re-think/become more aware of the consequences of our (intended/unintended) use of statistics in important areas as such. Thanks for sharing! In China, we have a saying, "众人拾柴火焰高", a metaphor that emphasizes the significance of making collective efforts as individuals. I wonder if we were to reflect on what we can start doing as individuals from the bottom-up perspective, as anyone who may or may not be competent enough to interpret a reported statistical outcome accurately, what are the ways we can use to reach more community members to help become critical readers of statistical outcomes in order to make informed decisions? For example, one recent phenomena I have observed is that although some people read news about the scientific reports (including the statistic) related to the seriousness/consequences of Covid 19 and the updated information on inflected cases, they were still skeptical about the trustworthiness of the reports or made poor decisions to act. What could we do about it as educators/individuals if we would like to intervene for the common good? Thanks!:-)