Rejected Hypotheses - local authority contamination

Following a flurry of media disparagement around the 2014/15 stats on local authority recycling rejects, Eunomia consultants Peter Jones and Andy Grant analyse England’s recycling data to truly understand the figures

Is the recently reported increase in recycling rejects in England (due to an increase in the amount of contamination in recycling bins) caused by growing confusion amongst the public, as the national press would have us believe? The Defra stats obtained by a BBC freedom of information request – and reported on with such disparagement by the likes of the Daily Mailand Daily Express – do show a substantial increase in the reported amount of local authority rejects. But, while an 84 per cent increase since 2011/12, and a 2014/15 figure of 338,000 tonnes of rejected material might sound bad, it’s still barely three per cent of the 11 million tonnes of material collected for recycling. For a complex, diversified system that relies on public participation, and which has to deliver material good enough to reprocess, a 97 per cent success is quite an impressive result.

Earlier this year, in contrast to the newspapers’ assumption that the increase in process rejects must be increasing public confusion, we identified a couple of other possibilities that would also explain the 184,000 tonne rise documented in Defra’s figures: better data capture, or better sorting at materials recovery facilities (MRFs). However, we didn’t attempt to assess which is the most plausible. In this article, we begin that tricky task.

We should also state a background assumption: any authority that collects a large proportion of its recycling co-mingled should, if it is reporting accurately, have a reject rate above zero per cent. A plausible minimum level of rejects is a matter for debate, but we think five per cent would be low and two per cent exceptional.


Rejected Hypotheses - local authority contamination

Kerbside-sort systems, though, can plausibly give rise to very low levels of rejects. Indeed, under the outgoing WasteDataFlow reporting rules, reject counts were focused on primary MRFs rather than any subsequent sorting by reprocessors. This made it possible for some multi-stream authorities to correctly report zero per cent rejects, although this would have overstated the amount of material put to beneficial use.

Systematic analysis

The 51 authorities that are currently using kerbside sort-type systems have an overall reject rate of less than one per cent. Comparing the 2011/12 and 2014/15 statistics, their reject rate has gone up by some 0.2 per cent, with rejects rising from 5,027 to 13,872 tonnes; but the data reveals that most of this increase is accounted for by issues with waste wood and kerbside recycling in South Gloucestershire. Since the reject figures are consistently and explicably low for these authorities, we have excluded them from the analysis that follows.

Part of the explanation for the overall increase in rejects might be the increasing popularity of single-stream and two-stream collection systems. Over the period since 2011/12, Eunomia’s records indicate that more than 60 authorities have moved from multi-stream to single-stream or twin-stream systems – the exact number depends on how you count authorities in waste partnerships. Of these, 23 reported zero rejects in 2011/12; only six did in 2014/15.

Rejected Hypotheses - local authority contamination

The total amount of reject material from councils that switched away from multi-stream collections accounts for 38,500 tonnes (13 per cent) of the increase in rejects between the two years. But a couple of observations diminish the significance of ‘switching’ as a contributing factor to the rise in rejects.

The rise in rejects in these ‘switching’ authorities is very similar to the overall average increase in rejects (about 575 tonnes per council).

More than 15,000 tonnes of the switchers’ increase was garden waste rather than dry recycling, and 14,300 tonnes of this was reported by the Dorset Waste Partnership. Excluding this outlier figure brings the average increase amongst ‘switching’ authorities significantly below the national average increase in rejects.

MRFs have indicated that they tend to look favourably on material from councils that have recently switched from a kerbside sort to a co-mingled system. The data perhaps supports the idea that such councils’ recycling yields relatively low levels of rejects at least in the early years after the system change.

Because of the very small contribution to rejects from authorities that switched away from kerbside sort, we also exclude these figures from the analysis that follows and focus simply on the collection authorities and unitaries that collected  some or all of their material co-mingled throughout the period 2011/12 to 2014/15.

Small change

Of these authorities, 157 saw their reject tonnage go up, while 47 saw it stay the same or decrease. But before we conclude that this reflects a widespread increase in public confusion, it’s worth looking at where the increase in rejects has taken place.

The top chart shows that 43 fewer authorities reported 0-2 per cent rejects in 2014/15 than in 2011/12, while the number reporting 2-10 per cent increased by 33. Where no authority reported a reject rate above 13 per cent in 2011/12, 11 did in 2014/15.

Perhaps more significantly, the great majority of the tonnage increase in rejects is accounted for by councils that in 2011/12 reported 0-3 per cent rejects, while the tonnage attributable to councils with 2011/12 reject rates above 10 per cent has actually decreased, as shown in the bottom chart.

This article was taken from Issue 86

While the analysis above is not conclusive, it is suggestive. Around 13 per cent of the 184,000-tonne increase in rejects appears to be associated with authorities that switched from source-separated collection schemes that are more likely to separate out non-target material before they ever get into the system. Around eight per cent seems to be attributable to garden waste issues in Dorset in 2014/15. A few per cent here and there are attributable to specific issues at individual authorities (some of which already had high rejects in 2011/12), which were explained in the BBC article that brought this issue to the fore – Greenwich using an ‘out-of-date’ MRF, for example, resulting in a 3,000 tonne increase.

But most of the rest has occurred across a wide range of councils that formerly had improbably low rejects; many still have reject rates below five per cent. Meanwhile, the authorities that had the highest level of rejects in 2011/12 have (taken as a group) seen their reject tonnage decreased.

Confused story

This is difficult to reconcile with the ‘confusion’ explanation. Are we to understand that confusion has increased only in areas where formerly there was none, but where people were more confused in 2011 about what to recycle, matters have improved?

It is also a poor fit with one of the alternative theories we advanced: that the figures accurately report improved sorting of rejects by MRFs. In fact, the starting level of rejects in the councils where most of the increase has occurred was implausibly low.

It is, though, consistent with another theory – better reporting by councils (or by MRFs to councils) of an issue that has been there in co-mingled systems all along.

This article is a version of one that originally appeared on the Isonomia blog. For more, visit 


Related Articles