Almost every article about discarded electronics opens with a number. Millions of tonnes generated annually. A single-digit or low-double-digit percentage properly recycled. A projected figure for some year in the near future. These numbers get repeated widely, and they are useful for conveying scale.
They are also considerably softer than their presentation suggests. The methodologies behind them vary, the definitions differ between reporting bodies, and the categories being counted are not always the ones a reader would assume. None of that makes the numbers wrong, but it does mean they support less precision than they appear to.
Anyone using these figures to make a case, whether in a policy submission, a sustainability report, or an article about the e-waste environmental impact, benefits from knowing what sits underneath them and where the uncertainty concentrates.
The Problem of Defining the Category
The first difficulty is that there is no universal agreement on what counts.
Some frameworks include large appliances such as refrigerators and washing machines, which are heavy and dominate any tonnage figure they appear in. Others cover only information and communication equipment. A statistic that includes appliances and one that does not are not comparable, and the distinction is often not stated.
Solar panels, electric vehicle batteries, medical equipment, and industrial control systems sit at the boundary and are treated inconsistently.
Cables, chargers, and accessories are sometimes counted separately and sometimes not counted at all.
This means that comparing figures from two sources frequently compares different things, and a change over time within one source may reflect a definitional change rather than a real shift.
How Generation Is Estimated
The tonnage generated in a year is not measured. It is modelled.
The usual method starts from sales data, applies an assumed average lifespan for each product category, and infers how much reached end of life in a given year. It is a reasonable approach and it depends heavily on the lifespan assumptions.
Those assumptions are difficult to validate. Devices are stored for years before disposal, passed to secondary users, or exported while still functional, and each of those behaviours shifts the timing of when a device becomes waste. The drawer full of old phones is a real phenomenon that models handle only approximately.
Sales data itself is incomplete in markets with substantial informal or grey imports.
The resulting figures are best read as indicating magnitude and direction rather than as precise quantities.
What Recycled Actually Counts
This is where the most significant ambiguity sits, and it is the number most often quoted.
Collection rate and recycling rate are different measures, and they are frequently conflated. Material entering a collection system has not necessarily been recycled; it has been collected.
Even within recycling, the measurement point varies. Some figures count material entering a processing facility. Others count recovered output. The difference is substantial, because not all input becomes recovered material.
Energy recovery, meaning controlled incineration with heat capture, is counted as recovery in some frameworks and excluded in others.
Reuse is treated inconsistently. A refurbished device returned to service is environmentally the best outcome available, and depending on the framework it may be counted as recycling, counted separately, or not counted at all. This has the perverse effect of making the best outcome invisible in some statistics.
And material handled outside formal systems, which is a large share globally, is by definition not captured in formal reporting.
Why the Documented Rate Is Probably an Undercount
Taken together, these factors suggest the commonly quoted low recycling percentages understate what actually happens to material, without changing the conclusion that a great deal is handled poorly.
Business-to-business flows are frequently under-reported. Corporate equipment sold into the secondary market, harvested for parts, or processed through commercial channels often does not appear in statistics built around municipal collection.
Informal recovery, whatever its environmental record, does recover material. It is not captured in formal figures.
Reuse, as noted, may be excluded from the numerator entirely.
The result is that the true picture is better than the headline figure in terms of material recovered, and the harm associated with how some of that recovery happens is not reduced by that fact.
Using the Numbers Honestly
A few habits keep the statistics useful rather than misleading.
State the source and the year, since figures circulate for a long time after the underlying data was collected and are frequently quoted as current when they are not.
Check what the category includes, particularly whether large appliances are in scope, before comparing two figures.
Distinguish collection from recovery, and say which one a number describes.
Treat projections with appropriate caution, since they extend models whose assumptions are already uncertain.
And resist the temptation to attach false precision to a modelled estimate. A figure given to one decimal place implies a measurement accuracy that the underlying method cannot support.
The broad picture the statistics describe is sound: the volume is large, it is growing, and a substantial share is handled in ways that cause harm. That conclusion does not depend on any particular figure being exact, which is fortunate, because none of them are.
