What is the e-waste problem?

I study the geography of electronic discards, often called 'e-waste'. I am interested in many aspects of it and try to approach the issue as what Latour (2005) calls a 'matter of concern'. There are many things at stake with e-waste as a matter of concern, but one of the important ones is 'numbers': whose numbers are correct? Indeed, the quantity, quality, and- more often -absence of data on or for e-waste is a source of controversy for those that study it. With that brief introduction, here are excerpts of a recent news item that offers a nice way into the controversy about e-waste and numbers (I have highlighted actors and sites with green and various claims in blue):

U.S. Isn’t Flooding the Third World With E-Waste By Adam Minter May 26, 2013 7:30 PM GMT-0230

Every year, Americans toss out as much as 4.5 million tons of old mobile phones, laptops, televisions, Xboxes and other electronic gadgets.

Some is recycled; some is repaired and refurbished for reuse; and some is thrown into landfills or incinerators. Almost none of it, however, is “dumped” overseas.

That, at least, is the conclusion of the first comprehensive survey of what happens to U.S. e-waste after it is dropped into a recycling bin. Published in February, the study by the U.S. International Trade Commission surveyed 5,200 businesses involved in the e-waste industry (companies that received the survey were required by law to complete it, and to do so accurately), and found
that almost 83 percent of what was put into American recycling bins in
2011 was repaired, dismantled or recycled domestically.

According to the same survey, only 0.13 percent of the 4.4 million tons of e-waste that Americans generated in 2011 was sent overseas for “final disposal” - a term that explicitly excludes recycling and reuse - with an additional --3 percent sent abroad for “unknown” purposes.

Reality is a far cry from the
long-standing claim, first made by the Basel Action Network, a
Seattle-based nongovernmental organization in 2002, that as much as 80
percent of U.S. e-waste is exported to the developing world. Amazingly, even with the wide currency the claim has enjoyed over the years among environmental organizations and the media, it was never based on a systematic study.

Misguided Efforts

This misunderstanding has led to several efforts at erecting partial export bans on U.S. electronics to developing countries, which - -other studies demonstrate -- import them as cheap and sustainable alternatives to new equipment. As a result, perfectly usable electronics are diverted into a recycling stream, where they are turned into raw materials, rather than into markets where they can be reused for years.

There are no statistics on how many used gadgets were exported from the U.S. to the developing world in 2002. Nor, for that matter, can anyone say for sure what happened to those gadgets. No doubt, many were broken down in developing-world facilities, where low-technology and often-hazardous methods of recycling and disposal were employed (such as the use of acids to strip copper and other metals from circuit boards in open, unprotected environments).

Anecdotally, I have been told by recyclers in southern China that cheap, secondhand electronics exported from the U.S. and, to a lesser extent, the European Union were used by Chinese computer labs, offices and dormitories in the 1990s through the mid-2000s, when new gadgetry simply wasn’t affordable. (There has been no comprehensive survey to verify these claims, however.)

It was a good deal for the U.S., too: In the 1990s and early 2000s, America didn’t really have an electronics-recycling sector, and those machines would have been put in a landfill if China hadn’t wanted them. Nonetheless, as China developed, and incomes rose, demand for those used machines dropped off.

The good news is that a similar cycle is occurring in Africa, where used electronics from the EU and the U.S. have become a critical means of bridging the global digital gap. Unlike Chinese imports in the 1990s and early 2000s, the African imports are being surveyed and quantified.

For example, a 2011 study by the United Nations Environment Program determined that only 9 percent of the used electronics imported by Nigeria - a country that is regularly depicted as a dumping ground for foreign e-waste - didn’t work or were unrepairable, and thus bound for a recycler or a dump. The other 91 percent were reusable and bound for consumers who couldn’t afford new products.

Initial Comments

In the above article excerpt I have highlighted actors and sites with green and various claims in blue. What do we learn as a consequence? First, that there are many actors and many sites in the issue of e-waste.

Actors and sites (in order of appearance):

the US
the Third World
Adam Minter (a journalist)
Americans
landfills
dumps
a survey
US e-waste
the US International Trade Commission
some 5,200 businesses (presumably US based)
a claim about how much e-waste the US exports to developing countries (since that claim enjoys wide currency)
environmental organizations
the media
US electronics
developing countries
other studies
low-technology recycling and disposal methods
acids
recyclers
European Union
Chinese labs, dorms, and offices
Africa
used electronics
Chinese imports
African imports
United Nations Environmental Programme
Nigeria
consumers

Second, we learn that the journalist, Adam Minter, claims there is a dispute or controversy: on the one hand, an environmental organization -the Basel Action Network (BAN) -claims that "as much as 80 percent" of US e-waste is exported to "the developing world". On the other hand, there are claims based on survey research that almost exactly match BAN's claim, but in reverse -"83 percent of what was put into American recycling bins in 2011 was repaired, dismantled or recycled domestically". So which number is correct - and is this even the right question to be asking?

Let's start with BAN's claim. In 2002 BAN published a report called Exporting Harm: the high-tech trashing of Asia. On page 1 of that report the organization states: "...informed recycling industry sources estimate that between 50 to 80 percent of the E-waste collected for recycling in the western U.S. are not recycled domestically" (BAN, 2002). Reading through the report several different versions of the latter statement are noticeable. For example, on page 11 the report reads, "[i]nformed industry insiders have indicated that around 80% of what comes through their doors will be exported to Asia, and 90% of that has been destined for China" (BAN, 2002); and on p. 14 under the heading 'How much E-waste is Exported?' the report states, "The short answer is that nobody really knows. yet anecdotal evidence on E-waste exported by the U.S. to Asia is abundant" and buttresses that claim with a photograph of asset tags accompanied a caption that reads, "Just some of many institutional labels from the United States found on computers in Guiyu, China in December 2001". The report continues with a statement that "[v]ery knowledgeable and informed industry sources, however, have estimated that around 80% of what is diverted to recycling [in the US] is actually exported to Asia" (BAN, 2002: 14). An endnote following the latter claim attributes it to a telephone interview BAN conducted with someone named Mike Magliaro of Life-Cycle Business Partners in Salem, New Hampshire in February 2002 (BAN, 2002: 14 and 50, endnote 41).

An initial analytical cut through BAN's claim tells us the following:

1. That BAN attributes the statistic of 50-80 percent to 'informed industry insiders' and 'very knowledgeable and informed industry sources', yet does not state who the sources are except in one instance (a phone interview with Mike Magliaro). Notice a very important shift that occurs within BAN's report then: a single attributed source is pluralized. The one becomes the several. This is a move that suggests there are many experts that make the same estimate about e-waste exports.
2. The photograph of asset tags that appears on p. 14 contains three such tags: one from the City of Los Angeles, one from the State of California, and one from the L.A. Unified School District. Together with the caption, the report implies that these tags are a mere selection out of an undisclosed number of such tags collected and photographed by BAN. So the photograph and the caption, work in combination to expand or pluralize the evidence said to support the 50-80 percent estimate. Thus, as in the previous point, three instances are made more plural. At the same time the report gives no information about how the asset tags were collected that would let a reader know whether those that are pictured on p. 14 come from some sort of systematic sample or not.
   

There is more to be said about the BAN report and the trace it leaves in the e-waste literature. I will return to that momentarily, but for now let's turn to Minter's article and its use of the US International Trade Commission (USITC, 2013) report to examine how truth claims are established in each.

As reported in Minter's (2013) article, the USITC's report is referenced to lend credence to the claim that "almost 83 percent of what was put into American recycling bins in 2011 was repaired, dismantled or recycled domestically". Minter does not quote the report, nor give a page number for the figure of 83 percent. However, when I search USITC's 250 page document the only reference to a figure of 83 percent that I can find occurs in a footnote on p. 4-10. The sentence and footnote read as follows: "Consistent with their primary activity, refurbishers accounted for the largest portion of UEP [Used Electronic Products] exports related to refurbishment and repair in 2011 - 72 percent by volume, or 60.2 million units.³²" (USITC, 2013). Footnote 32 reads, "Low-precision estimate, with an RSE [Relative Standard Error] of 83 percent" (USTIC, 2013: 4-10). Now, the figure described in footnote 32 is interesting. There the reference to 83 percent is to a measure of statistical error, not to the a measure of what proportion of UEPs were recycled domestically in the US. An RSE of 83 percent is very high. The lower an RSE is, the more precise a result is statistically speaking (the USITC report gives a nice example in Box 1.2 on p. 1-8). However, in the 'Key Findings and Observations' section of the USITC report, we learn that exports of UEPs from the US amount to "7 percent of total sales" of UEPs (USITC, 2013: xiii). The corollary of the 7 percent export figure is, by straight arithmetic, that 93 percent of total  sales of UEPs are domestic to the US (a figure confirmed by data given in Table ES.1 on p. xiii of USITC, 2013). So whiter the figure of 83 percent? The following seem possible:

1. Minter mistook a measure of statistical error about exports for a measure of domestic recycling. This possibility seems unlikely given the other possibilities below.
   
2. Minter miscalculated the 83 percent as a corollary to the 7 percent figure (if so and the USITC figures are correct, then the picture is even rosier - almost 93 percent of 'what was put into American recycling bins in 2011 was repaired, dismantled or recycled domestically')
3. Minter has a source related to USITC (e.g., one or more of the authors of the report) who gave him the 83 percent figure
4. Minter has a source for the 83 percent other than USITC and is misattributing it to that report.
   

If the first possibility is as unlikely as I think it is, then it would seem that, according to Minter and USITC, somewhere between 83-93 percent of "of what was put into American recycling bins in 2011 was repaired, dismantled or recycled domestically" (Minter, 2013).  We are now confronted with a range more or less diametrically opposed to the range supported by BAN (50-80 percent exported).

In contrast to the BAN report, the USITC document is at pains to demonstrate its methodological rigour. It has a chapter section (p. 1-7 --- 1-11) and at least four appendices (Appendices A, E, F, and H) devoted to describing its methods. These methods include a survey of businesses in the UEP business, the use of both public and confidential census data on exports of a variety of electronic products, a public hearing, visits to various types of facilities in the UEP sector (e.g., recycling plants, asset disposition firms, original equipment manufacturers), and a literature review that included BAN's (2002) report. All of these methodological approaches have their pros and cons and the report makes at least some of these explicit. It notes, for example, that several participants in the public hearing voiced concerns about a survey approach since they found it unlikely that respondents would accurately report export activity that might raise questions about its legality (see USITC, 2013: 1-8, footnote 28). Whatever the various shortcomings of these methods may be, the move made by the report is to make those shortcoming public and, in so doing, enhance the truth claims of the report.

Taken together, then, we can recognize forms of writing - even literary techniques or rhetoric (in the positive sense of that term) - at play in all of these documents: the BAN report, Minter's article, and the USITC report. In this sense, they are all the same. Their use of numbers all rely on modes of representing those numbers in such a way as to bolster their claims to trustworthy knowledge. This literary technique is what science and technology studies (STS) scholars like to call 'modest witnessing' (see for example Shapin and Shaffer, 1987; Haraway, 1991). In short, modest witnessing is a literary technique in which the author(s) of a truth claim present that claim in such a way as to suggest that the claim would be judged true by anyone who followed the same methods as the author(s) and, more importantly, that the claim so judged does not come from the author(s) but from some other authoritative source(s) such as instruments and/or acknowledged experts (STS scholarship has a long critical but also celebratory engagement with modest witnessing; see Shapin and Schaffer, 1987 and Haraway, 1991 for a taste).

Given that all the documents discussed so far rely on the same rhetorical moves, it begs the question of why there is a dispute, a controversy over numbers about e-waste. Let's trace the influence that these documents have had in terms of generating statistical utterances about e-waste.