SAJEMS NS Vol I (1998) No 3 422
An Economic Case 'for Regulating Against the
Use of Non-Deposit Carrying Glass Containers
of Beverage in South Africa
SHosking
Department of Economics, University of Port Elizabeth
ABSTRACT
As a result of minimal private cost many people dispose of non-deposit bearing
glass containers in ways which cause glass pollution: hazardous broken bottles
and litter. This pollution imposes costs on users of the affected environment
and on municipalities, which have most of the responsibility to clean up,
although in South Africa the two main glass packaging producers also play a
role by operating a recycling system. A case study was carried out in the Port
Elizabeth area in which exploration is made of the glass that does not get
recycled and an intuitive analysis is made of the costs of different options for
managing recyclable glass waste. It is concluded that the case deserves further
investigation for introducing legislation in South Africa making bottle deposits
mandatory.
JELK200
INTRODUcnON
From an environmental perspective one of the most promising ways to deal with
glass packaging pollution is to recycle it. If reprocessed and used again,
alcoholic and soft drink container bottles do not deface the landscape or add to
the mound of refuse generated by the modem industrial 'throwaway' society -
ultimately requiring burial in municipal tips. Fortunately, under some
circumstances it is profitable to recycle glass bottles, and as a result, a recycling
glass bottle industry has evolved.
Despite the development of this industry, the problem of glass bottle waste still
remains an acute environmental problem in South Africa. For this reason the
question arises of whether the government should not step in and help remedy a
problem the market is failing to, by introducing more regulation in the beverage
bottling industry.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
423 SATEB NR Vol I (1998) Nr 3
It is not an option to be adopted on weak grounds, because this type of
regulation can undermine economic freedom (Barry, 1994': 6). Specifically, if
the government legislated a mandatory deposit-refund system covering all soft
and alcoholic drinks, the freedom to enter into contracts for beverages contained
in non-refundable glass bottles would be lost. This paper will explore the
economic case for this option, using Port Elizabeth as a case study
THE NATURE OF THE PROBLEM
This type of regulation is not new in the world. It is usually called container
deposit legislation elsewhere. In the United States discussion of container
deposit schemes can be traced back to the 1930s (W.D. Scott and Company Pty.
Ltd., AUgust, 1983: 10). The legislation encompasses containers of all types of
beverage. (The type of regulation being considered in this report merely relates
to glass.)
The following states of the United States of America had introduced bottle bill
(container deposit) legislation by 1989: California, Connecticut, Delaware.
Iowa, Maine, Massachusetts, Michigan, New York, Oregon and Vermont
(Tietenberg, 1992: 206). There is similar legislation in Canada (Turner, Pearce
and Bateman, 1993: 259-260). In Australia, South Australia introduced
container deposit legislation in 1977 (Business Regulation Review Unit, June,
1989: 11 ). There has also been strong support for various aspects of container
deposit legislation on the continent of Europe. The following European
countries have product charge or deposit-refund legislation covering beverage
containers: Austria, Sweden, Finland, Gennany, Norway, Switzerland and
Denmark (Turner, Pearce and Bateman, 1993: 259-260).
In South Africa legal protection of the environment can only be justified if it is
in human interests; the legal principle being, cum igitur hominum causa ius
constitulio sit (law is an institution for the sake of men) (Rabie, 1973:1).
Currently, legal protection against careless disposal of glass bottles is justified
in terms of this legal principle on the grounds that, like other littering, it is an
eye-sore and or health risk for other human users of the environment. The
protection takes the form of litter deterring provisions in the Environmental
Conservation Act (No. 73) of 1989. In terms of Section 19 of the Act a fme of
Rl 000 and/or 3 months imprisonment may be imposed on a person for littering
a public attraction. In sections 20-24, regulations and directions of waste
management are dealt with. They provide for the imposition of fmes of up to
Rloo 000 and/or 10 years imprisonment on people who dispose of waste
illegally. Is this enough, or should the government be doing more?
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol I (1998) No 3 424
Of the representatives spoken to in the beverage industry in S<1uth Africa, few
were prepared to commit themselves on the desirability of container deposit
legislation in South Africa. Most representatives spoken to did not foresee it as
likely in the near future and redirected inquiries on the issue to Owen Bruyns,
Executive Director, Packaging Council of South Africa (pAC SA),
Johannesburg. Telephonic and face to face discussions were held with Mr.
Bruyns on the topic during second half of 1996. He is of the view that the case
for regulating against non-returnable bottles is weak because:
• regulation is unnecessary • the bottle manufacturers are already
addressing the problem, by administering non-profit making glass
recycling systems (e.g., bottle banks),
• the impact of the proposed regulation has been researched in other
countries (USA and Australia) and shown to be an inefficient solution to
the problem of glass litter,
• it will not solve the relevant problem but merely change the nature of it -
the relevant consumers will simply switch from disposable glass
containers to disposable tinplated steel and aluminum cans and to
disposable plastic PET bottles,
• it will reduce consumer surplus by reducing choice and pushing up the
prices of the relevant beverages (because they will be more costly to
produce), and
• it will reduce jobs and income (and thereby tax receipts) in the glass
bottling and beverage industries and/or divert capital to the distribution
and cleaning sectors.
The first of these arguments is the one primarily addressed in this paper. The
second is addressed in sections 2 and 3, but the following issues are not
addressed:
• possible consumer substitutions of glass for plastic, aluminum and
tinplated containers,
• possible changes to beverage prices, and
• possible changes to employment and income totals in different industries,
e.g., due to reductions in some types of bottle production and increases in
other, and to increases in recycling activities.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
425 SATEB NR Vol I (1998) Nr 3
FINDINGS OF SELECTED OTHER STUDIES INTO THE ECONOMICS
OF LEGISLATING AGAINST NON-RETURNABLE BOTTLES
The studies to which PACSA refer are those on container deposit legislation by
W.O. Scott and Company Pty. Ltd. (August, 1983) and the Business Regulation
Review Unit (June, 1989) - both relate to the case of Australia. They are
discussed under (a) and (b) below.
(a) W.O. Scott and Company Pty. Ltd. (August, 1983) analyzed the probable
economic impact of introducing container deposit legislation throughout
Australia and concluded that:
• it would increase the cost of soft drinks and beer in cans and plastic
containers, and for this reason also their retail prices, as well as the
retail price of drinks in multi fill containers (to help cover these
increased costs),
• skilled or semi-skilled manufacturing jobs would be replaced by
unskilled bottle sorting and handling jobs, with the result that
employee earnings would decline,
• government revenue would decline because sales would decline,
profits would decline, employee income would decline and
production plants would be written off, as unprofitable plants were
closed down, e.g., tinplated steel can producers,
• overall energy consumption would decrease, but more scarce primary
energy sources would be used due to increased transport requirements
(less hydroelectricity, coal and natural gas, but more oil),
• the raw materials used in the beverage container industry of Australia
are abundant in the country,
• the experience of South Australia with mandatory deposit legislation
was that the reduction in costs of litter management were insignificant
- only a 3% reduction in total litter, and
• alternatives to mandatory deposits are more cost effective in reducing
litter and conserving resources than mandatory deposits, e.g.,
improving return rates on existing multifill packages, improving
recycling rates and funding comprehensive litter reduction campaigns.
(b) The Business Regulation Review Unit (June 1989) considered the same
topic as W.O. Scott and Company Pty. Ltd. (August, 1983), but a few
years later. It noted that between 1973 and 1978 counts of litter fell by
80% in the state of South Australia and that container deposit legislation
was introduced there in 1977. In South Australia mandatory deposits
apply to all beer, wine cooler, soft drink and milk containers. The
Business Regulation Review Unit were not of the view that this reduction
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol 1 (1998) No 3 426
was purely and simply due to the introduction of the relevant legislation,
because some years prior to 1977 there had been vigorous anti-litter
media campaigns in the state and the introduction of on-the-spot fines for
littering. Moreover, impressive reductions in litter also occurred in states
where there was no container deposit legislation, e.g., 63% in New South
Wales between 1979 and 1985 and 50% in Victoria between 1983 and
1988.
Some of the negative features of container deposit legislation were:
• it imposes a particularly high deposit on cans, which cannot be re-
used, thereby raising the overall price of the product and reducing
demand,
• it increases production and retailing costs (bottle washing plants,
storage of bottles at retailers, slower bottle filling compared to cans
and increased distribution costs),
• it increases consumer costs by making them return bottles, and
• it increases the cost of market entry (e.g., beer from Western
Australia) thereby facilitating local monopolies.
The conclusion of the Business Regulation Review Unit was that the cost
savings, brought about by container deposit legislation as a result of
reduced garbage disposed, were far outweighed by the increased costs
this legislation imposed on consumers and producers. The Unit
suggested that more efficient ways of reducing costs would be privatizing
waste disposal operations, media campaigns, increased littering fines and
appealing to beverage and container producers to avoid ring tops (on
cans) and other devices that can add to the problem of litter.
One of the well known cost benefit studies carried out on bottle deposit
legislation in the United States is that of Porter's (1978).
(c) Porter (1978) conducted a cost benefit study of a proposal to legislate, in
the state of Michigan in the USA, for the introduction of mandatory
deposits on beer and soft drinks containers. The proposal was to make
all beverage containers sold in the state of Michigan subject to a
returnable deposit of 10 cents. At the time only 27% of the containers
were returnable. His conclusion was that it depended upon the relative
valuations of the average citizen's time sacrifice cost and willingness to
pay for a reduction in container litter. For instance, he calculated that a
desirability of mandatory deposits equated to an aggregate willingness to
pay about $200 million per annum (by the 9,1 million citizens), or $27.26
per annum per citizen (assuming the average citizen's time return cost
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
427 SATEB NR Vol 1 (l998)Nr3
were equal to the 10 cents deposit proposed). Porter suggests that this
sum may be higher than what the citizens would be prepared to pay for a
reduction in container litter. However, it is noted that his justification for
a return cost estimate equal to 10 cents per returnable bottle is based on a
rather extreme assumption; that the return cost of the non-returnable
bottle is trivial and can be ignored. What Porter's analysis strongly
suggests is that no general answer may be given to the question of
whether mandatory deposits are desirable.
(d) Based on the findings of a cross-section of cost benefit studies done on
deposit-refund systems in Europe and the United States, Turner, Pearce
and Bateman (1993: 264-265) conclude that they do not generate
significant net social gains. The problem, as they see it, is that they do
not reduce the overall volume of municipal waste enough to offset the
additional administrative costs.
(e) Although little support has been found amongst industrial groups in the
United States for government legislated deposit-refund systems, the
general public have been very supportive of them (Turner, Pearce and
Bateman ,1993:255). There is evidence that the public respond
positively to changes in the pricing of waste disposal. The examples of
Highbridge and Seattle show how changing the incentives to dispose of
litter have worked in the United States (Tietenberg, 1992).
• In January, 1988, the town of Highbridge in New Jersey, USA,
replaced its flat 280 dollar annual fee for the collection of refuse with
a fee that varied with the amount of trash (New York Imles. 24
November, 1988, BI,B7). Each household was given 52 stickers for
140 dollars. Additional stickers could be bought at 1.25 dollars each.
As a result peoples attitudes towards wastes changed rapidly: the
amount of refuse required to be collected went down 25% as residents
began to compost food and yard wastes and recycling programmes
flourished in glass, newspapers and cans. In 1989 the base sticker
rate was increased to 200 dollars and additional stickers to 1.65
dollars each. This increase gave rise to concerns that the new pricing
strategy for refuse removal was going to place a greater burden on the
poor. However, that concern was apparently misplaced. Under the
old system every household paid the same fee for refuse collection,
regardless of how much refuse was put out, and as poorer households
were producing less refuse, they were in effect subsidising the
wealthier ones.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol I (1998) No 3 428
• Seattle. Washington instituted a similar system. but with higher
prices. They too experienced a dramatic decline in refuse disposal
and increase in recycling.
The lesson learned is that when the incentive not to waste is made direct.
people change their behavior and waste less. The refund system is similar
in its effecL A deposit must be paid by the consumer, on the purchase of
the recyclable item. which reflects the cost of recycling or disposing of iL
Glass bottles and aluminum cans are good examples of such items. but
there are also many others to which the system has been successfully
applied. including cars and car parts (Bohm. 1981. pp.120-124). On
returning the item to a designated dealer a refund is given. This refund
creates the incentive for someone. not necessarily the purchaser. to collect
the item and transport it to the designated dealer. If the purchaser
considers the incentive insufficient to return it. an opportunity for others.
probably poorer people. is created to make money out of the system by
returning it.
THE MARKET DETERMINED EFFICIENT LEVEL OF GLASS
RECYCLING
Investigations of the literature on recycling confinn that in an efficient market.
devoid of any imperfections, there is no need for intervention because the
interaction of supply and demand bring about an efficient level of recycling
(Tietenberg, 1992). In this kind of market the rising costs of virgin materials
and of waste disposal increase the attractiveness of recycling from both the
demand and supply sides of the market. Products relying exclusively on virgin
raw materials would be subject to higher prices than those drawing on recycled
materials. Consequently, consumers would switch from products relying
exclusively on virgin raw materials to ones drawing on recycled materials; a
switch known as the composition of demand effect (Tietenberg, 1992).
The theory that there is an efficient level of recycling glass waste in South
Africa does not relate to all glass but only to that which is potentially profitable
to recycle; mainly to glass used as packaging (bottles and jars).
There are many private economic incentives to recycle glass. Glass containers
of beverage are almost 1 WIG recyclable. which means they can be recycled with
minimal loss of material, or emission of toxic waste. over and over again.
Moreover, recycled glass melts at lower temperatures than the raw materials
used for glass production. For each 10% increase in recycled glass containers
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
429 SATEB NR Vol 1 (1998) Nr 3
added to the mix, the melting en~ required is reduced by about 2,5%. As a
result both energy usage and wear and tear on capital (the t\imaces) are reduced;
two of the main costs incurred in producing glass (Glass Recycling Association,
November, 1996). In the furnaces, molten glass is heated up to a temperature
of between 1 260 and 1 480 oC.
Notwithstanding these energy and capital saving incentives, there are limits to
the amount of glass that it is profitable to recycle. The main limiting factors are
the collection, processing and transport costs. At some point the energy and
capital saving of using recyclable glass is outweighed by the greater acquisition
cost of cullet (crushed used glass packaging) over that of the virgin raw
materials used in the glass mix (pers. comm., Owen Bruyns, 1996). It is a point
which is reached fairly quickly in South Africa, because the virgin raw
materials used to make glass are plentiful: silica sand (58,1% of the mix
required), soda ash (18,2%), limestone (15,6%), feldspar (4,4%) and dolomite
(3,7%). There is also a quality deterioration problem (increased bubbles in the
finished products) if more than 40% of the input material used to make glass is
made up of cullet (Vogler, 1981:173). This limiting factor is also relevant in
Australia (Business Regulation Review Unit, June, 1989: 35).
Of potentially recyclable glass about 65% recycles in South Africa: about 43%
in the form of returnable bottles and about 22% through bottle bank and buy
back systems set up by the Glass Recycling Association (Glass Recycling
Association, November, 1996). Glass recycling with returnable bottles is
generally found to be more cost effective than with non-returnable ones
(Vogler, 1981: 178).
It is deduced that the market choice in South Africa is to recycle 65% of
packaging glass. If efficient markets are assumed it is deduced that of the
potentially recyclable glass 35% is inefficient to recycle (the profit incentive is
insufficient).
RECYCLING BY SELECTED BOTILING COMPANIES IN SOUTH
AFRICA
Softdrinla
According to PACSA, as a proportion of the total volume packed in returnable
and non-returnable containers, the volume of soft drinks packed in returnable
(voluntary deposit) containers in South Africa increased from 65% to 67,4%
between 1990 and 1993. Of the returnable bottles in 1993: 54% were 1 litre
glass bottles, 7,3% were 1,5 litre plastic bottles and 6,1% smaller glass bottles.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol 1 (1998) No 3 430
Of the non-returnable containers: 65% were metal cans (as against 40% in
1990) and 27% plastic bottles. By comparison with the United States a much
higher proportion of soft drinks are packaged in returnable containers in South
Africa. In the United States this proportion is only about 8% (PAC SA, 1996).
The main bottling company of non-alcoholic beverage (Coca Cola) in South
Africa, the South African Bottling Company (SABCO), has its head office in
Port Elizabeth. In November 1996 it reported that 67% of the packaging done
by it was in glass containers, and 57,11% of the glass packaging was in
returnable containers (Ernie van Vuuren, Manager SABCO, November, 1996).
About 6% of the returnable bottles are damaged in the return process. Whereas
non-returnable bottles are merely rinsed, returnable ones first have to be sorted
(which is very labour intensive), the caps removed, and then washed. Dave
Davies, also of SABCO (production manager), is of the view that consumer
preferences are shifting from soft drinks packed in returnable containers to ones
packed in non-returnable containers, but no evidence was offered to support this
view (November 1996). The converse appears to be true in the beer market.
Beer
For the period, June 1986 to January 1994, South African Breweries reported to
PACSA that as a proportion of total sales of all malt beer in South Africa, the
market share of:
• returnable re-usable voluntary deposit containers increased from about
72% to 79010 (and the 750ml specifically, from 64% to 75%),
• non-returnable metal cans increased from 6% to 10%, and
• non-returnable dumpies decreased from 9% to 7%.
It appears that since 1993 the swing away from non-returnable to returnable
containers has continued in the beer market. In November, 1996, South African
Breweries Port Elizabeth office (Owen Kingwall, Production Manager,
November, 1996) reported that of its total sales, market shares by types of
container were roughly as follows:
• 80% in 750 ml 'quarts' (returnable),
• 10% in 375 ml 'pints' (returnable),
• 5% in 340 ml 'dumpies' (non-returnable), and
• 5% in cans (non-returnable).
A returnable beer bottle can be re-used up to 30 times per bottle (Owen
Kingwall, South African Breweries, Production Manager, November, 1996).
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
431 SATED NR Vol 1 (1998) Nr 3
RECYCLING IN PRACDCE IN SOUTH AFRICA
In South Africa the Glass Recycling Association administers glass recovery
systems. This association was fonned in 1986 by the main two glass packaging
producers in South Africa, Consol Glass and Metal Box Glass. The rationale
this association provides for its activities is 'to promote the concepts of a clean
environment and a healthy community' ('A Message From the Glass Recycling
Association', November 1996). It employs two methods of glass recovery: a
bottle bank system and the buy back system.
The bottle bank system works by consumers putting their unwanted glass jars
and bottles into one of about 1490 glass bottle banks set up in 11 S cities and
towns in South Africa, and agents of the company collecting this glass from
these banks.
Traffico is the agent appointed to administer the bottle banks in Port Elizabeth.
There are currently about S4 of them in the city. It collects about % ton of glass
per week from these S4 bottle banks (Willem Jordaan, Operations Manager,
Traffico, November 1996).
The buy back system is a repurchase arrangement with 120 appointed agents in
designated areas in South Africa. The arrangement is for the purchase by
Consol Glass of crushed used glass packaging. known as cullet. To qualifY as
cullet and be repurchased by these companies:
• only bottle and jar glass should be used
• glass should be separated by colour (into amber, green, flint or mixed),
and
• no contaminants should be mixed in amongst the jar and bottle glass - any
other type of material, including other types of glass (such as window,
light bulb, drinking and crystal), and
• the glass should be crushed (culled) and transported to either of the
Johannesburg or Cape Town Consol Glass plants.
In Port Elizabeth, Traffico sees to it that these requirements are met. The bottle
banks only make up part of its supply. SABCO supplied it with 29 tons of glass
in 1996; bottles damaged in the process of production (Ernie van Vuuren,
SABCO, January, 1997). 'Totting rights' to collect glass waste at the
municipal tip site in Port Elizabeth also yield it a supply (Ken Kendel. PE
Municipality, November, 1996). as does South African Breweries in the fonn of
their damaged bottles (Owen Kingwall. Production Manager, South African
Breweries, November. 1996).
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMSNS Vol I (1998) No 3 432
WHERE THE MARKET (AND CONSUMER MORALITY) FAILS
The motivation provided by the Glass Recycling Association for its recycling
activities is moral (rather than profit). It suggests implicit acceptance of a weak
. version of the producer pays principle by the two main glass packaging
producers in South Africa, i.e., acceptance that part of the disposal cost relating
to the non-returnable bottles is their responsibility. We may deduce that they
feel that the other part lies with the consumer, in the form of a moral obligation
to take the non-returnable bottle, from which they consume the contents, to one
of the many bottle banks they provide.
This position is a reasonable one in an era when the public is continuaHy urged
to be environmentally conscious, and environmental education and awareness
campaigns are regularly conducted.
However, in many circumstances, taking a bottle to the bottle bank is neither the
most efficient action to take, nor the one people will necessarily be inclined to
take. For a self-interested person, the most efficient action with respect to
non-deposit bearing glass containers is to discard them as soon as possible after
the consumption of their contents, because this imposes the lowest private cost
on them. Minimizing the return cost is undoubtedly the main reason why
people would want to purchase this type of bottle instead of the returnable one.
On privately owned land, this means in a general refuse bin, and on common or
government land, this means in the park, on the beach, alongside the walkway,
or on the street. For people who are more sensitive to the social benefits of
responsible behaviour with respect to litter, one would expect them to make the
effort to take their non-returnable bottles to the bottle banks. However, even for
these people, there are circumstances where they may not do this. For instance,
if they perceive large numbers of others about them not making the effort, they
may deduce that their actions are of so little consequence that they cease to
bother themselves (i.e., the public good 'assurance' problem).
For these reasons (despite public education and awareness campaigns) 35% of
the glass packaging industry's total output of about 125 000 tons per annum
ends up either being buried in landfills, together with other refuse or left on the
ground, frequently in broken pieces (Glass Recycling Association, November.
1996). In landfills the external cost is passed on to the municipal rate payers
(estimated at Rl5 million per annum for South Africa) and to refuse workers
who are injured in the process of transporting it pieces (Glass Recycling
Association, November. 1996). When left on the ground the external cost of
packaging glass waste is more difficult to estimate, but it may be even greater
than for the landfill case. Left broken on the ground it is time consuming to
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
433 SATEB NR Vol I (1998) Nr 3
pick up, it detracts from the visual appeal of the environment and is physically
dangerous to those barefoot. The problem is particularly evident at places
where picking up glass must be done by hand, litter spoils the look and feel of
the place, and where it is natural for people to want to walk barefoot, for
instance, the beach. Recreational values are reduced. Significantly, on many
beaches glass litter is simply left by the administering authorities, presumably
because the cost of cleaning up is prohibitively high (see Port Elizabeth case
study below).
To sum up: an important cost is left out of the calculus that underlies the market
outcomes on what is recycled in South Africa, namely, the cost that is incurred
in disposing of the 35% of glass produced per annum, not recycled. Currently
part of this external cost problem is bome by local governments, in the form of
refuse collection, cleaning operations and municipal tip management, and part
by the citizens living and using the environments where the glass packaging is
disposed.
The question this paper addresses is whether this external cost can be reduced
by making a deposit-refund system mandatory for glass beverage packaging,
without increasing other costs by more than what is saved.
ASPECfS OF GLASS WASTE DISPOSAL IN PORT ELIZABETH
Glass waste - a municipal perspective
The city of Port Elizabeth collects about 1 000 tons of solid waste per day from
households and firms (Ken Kendell, Chief Cleansing Officer, November 1996).
The costs of refuse collection amount to about &23,5 million. In addition about
R15,3 million is spent on cleaning up streams and beaches. The cleansing
division of the PE municipality employs about 730 people and spends more than
Rl million per week in total on cleaning.
A domestic refuse analysis conducted in 5 suburbs in Port Elizabeth
(Summerstrand, Hollard Park, Newton Park, Sydenham and Framesbury) for the
city of Port Elizabeth shows that glass accounts for about 7 % of the weight of
the refuse and 2% of the volume (Table 1).
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol I (1998) No 3 434
Table 1: Cumulative constituents analysis of 5 samples of domestic
refuse in Port Elizabetb.
Constituents Sample Weigbt Weigbt%of Volume 0/. of
(kes) lSJ!- Sample Total Paper 249,85 38,04
Glass 99,3 7,9 2,36
Metal 43 3,ll 3,97
Plastic 50,5 3,66 13,96
Veg. & Putrescible 497,55 36.04 11,62
Garden Refuse 373 27,02 25,85
Rags 13,5 0,98 1,41
Unclassified 29,6 2,14 0,68
Wood & Combustible 24,25 1,76 2,11
Totals 1380,55 100 100
Source: City of Port Elizabeth, ChiefCleansmg Officer, November 1996.
As a result of broken glass bottles being placed in black refuse bags, about 2
injuries per week are sustained by refuse collectors firms (Ken Kendell, Chief
Cleansing Officer, November, 1996). Mostly these injuries are superficial cuts,
but there also have been deep leg cuts and kidney damage reported. The
presence of glass is also dangerous for those who seek to collect specific items
from the refuse at the municipal tip site; which is the reason why the PE
municipality is currently considering installing an expensive a conveyer belt
system at the tip site.
Glass waste on the beaches
The highest concentration of Port Elizabeth's glass pollution is found on its
beaches and recreational areas, especially on its northern beaches, where the
highest numbers of beach-visitors are found (Ken Kendell, Chief Cleansing
Officer, November, 1996): On New Year's Day, 1994/5, about 132000 people
visited PEs northern beaches of Joorst Park, St. Georges, and Wells Estate, and
about 92 000 PEs southern beaches of Kings, Humewood, Hobie and Pollock
(95196 Summer Season Lifesaving Report, City of Port Elizabeth). Of the
incidents reported by the lifeguards at Humewood and Hobie beaches during the
1995196 Summer Season Lifesaving Report:
• 8 were for cuts,
• 39 for 'helpouts',
• 2 for sprains or fractures, and
• 3 for stings.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
435 SAlEB NR Vol 1 (1998) Nr 3
There were 10 injuries reported as being the result of broken glass by lifeguards
on the northern beaches of Port Elizabeth (1oorst park, Wells Estate and St.
Georges Strand); two of which were considered serious. The senior lifeguard
for the area (Armand) was of the view that excessive use of alcohol was a
contributing factor to the problem of glass litter on the beach. He deduced that
inebriated people did not to think or care about the health risk imposed on other
beach users resulting from the disposal of glass on the beach.
One of the most common types of glass litter found on the beaches of the Port
Elizabeth (PE) area is the non-returnable beer dumpie; mainly containers sold
by South African Breweries. The buyers of these containers constitute a niche
market for South African Breweries. 10% of the total market (Owen Kingwall,
South African Breweries, November 1996). These consumers are happy to pay
a bit more for their beer in order to avoid the time and transport cost of
returning the bottles after consuming their - presumably the more affluent
section of the market Beer bought in single-fill bottles is a bit more expensive
than that bought in returnable ones, after the deposit for the bottle has been
claimed· usually between 3 and 8%.
The problem of glass litter is also evident on other beaches near PE, but outside
PE municipal control. During 1996 Jean Spearpoint and David Brown
monitored solid waste litter on a 100 square metre section of the beach adjacent
to the Van Stadens Mouth Resort. The cleaning up of the beach is the
responsibility of the Department of Culture and Recreation, Eastern Cape, and
the Eastern Cape Conservation department. Access to the Van Staden Mouth
Resort and beach is controlled and takes the form of a R15 entry fee per car
levied by the Department of Culture and Recreation, Eastern Cape, at the
entrance to the resort. They recorded details of any litter they found in random
walks during the 6 months from May to October 1996. The results of this study
are presented in tables 2 to 4 below.
Table 2: Composition of pollution according to the material of items of
litter picked up in random walks May to October, 1996, on a
100 metre stretch of beach at Van Stadens Gorge.
Material littered Number of items % of total
Glass 507 86
Metal cans 76 13
Plastic 6 1
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol 1 (1998) No 3 436
Table 3: Number of units of glass and weight of broken glass picked up
in random walks May to October 1996, on a 100 metre stretcb
of beach at Van Stadens Gorge.
Month in 1996 Number of Units Weight of broken
glass in Kgs
May 53 1,82
June 67 5,45
July 94 2,84
August 131 1.14
September 114 3,64
October 48 2,39
Total 507 17,28
Table 4: Proportion of items of glass pollution made up of deposit
bearing containers and non-deposit bearing.
%
7
93
The particularly interesting aspect of this research is that it indicates what litter
one could expect to come across on the beach after organized cleaning has taken
place. Glass is the litter one is most likely to come across and 93% is in the
form of non-deposit bearing glass containers.
Mike Bentall, Chief Environmental officer for the East London municipality,
reported that similar problems are encountered in East London's public places
(August 1996).
The public's view on the disposal of glass bottles in Port Elizabeth
During November 1996, H. Gous carried out a limited public survey of people's
views in Port Elizabeth on the disposal of glass bottles. The survey method
employed was personal interviews with people randomly selected from the more
affluent suburbs of Port Elizabeth. The selection of people from the more
affluent suburbs was motivated by the smail scale of survey proposed and the
observation that in the beer market, non-returnable bottles were targeted at the
wealthier consumers - poorer ones (the majority) preferring the returnable
containers because the beer is cheaper out of them. Had poorer people been
included in the survey we would have expected a far smaller proportion of the
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
437 SATEB NR Vol 1 (1998) Nr 3
sample to indicate that they purchased non-retmnable bottles. A total of 100
people, aged from 18 to 57 years, were interviewed.
Of those interviewed (l00):
• 96 knew which beverage containers were deposit bearing,
• 45 buy returnable g1ass bottles only, if the option exists,
• 85 purchase between 1-5 bottles of beverage per week, 10 purchase 6-10
bottles and 5 were unsure,
• 60 disposed of their non-returnable bottles in the bottle banks provided
by the Glass Recycling Association, and
• 35 regarded the use of bottle banks as too inconvenient.
M. Akkerman undertook a similar survey in the Northern Suburbs of Cape
Town during November, 1996, but asked some additional questions - ones
focusing on why people do not use the recycle systems currently available. He
interviewed 16 people. Of these people:
• all thought that the recycling of g1ass was a good idea in principle,
• 13 knew which beverage containers were deposit bearing and which not
(although some expressed confusion with respect to bottles containing
beverages other than soft drinks and beer),
• 10 regarded returning non-returnable bottles for recycling as
inconvenient, and
• 13 thought that they should be compensated monetarily for time and costs
incurred in returning bottles for recycling.
G/ass as hazardous waste
Much of the g1ass collected in Port Elizabeth is not recyclable (see Oliver, 1975,
on the making and using of glass other than for packaging). This glass is
disposed of by being treated with chemicals to immobilize heavy metals,
compacted and covered at a hazardous landfill site. Waste Tech gives
assistance in the process on request (Susan Alcock, Waste Tech, November,
1996) and also itself disposes of some hazardous g1ass waste. It removes glass
waste from Shatterpruff'e, ELMOSA and Associated Glassworks.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol I (1998) No 3 438
TOW ARBS A CASE FOR REGULATION
The object of Regulation
A model is provided in Figure I (below) where the aim of regulation is
described.
Figure 1: A Single Market Model of the Case for Regulation
()
% OF RECYCLABLE GLASS WASTE DISPOSED
INTO mE ENVIRONMENT
100
In Figure 1 two functions are shown: MCr and MC •. The negatively sloped one
(MCr) shows the marginal social cost of reducing the amount of glass packaging
disposed into the environment. The positively sloped one (MCi) shows the
marginal social cost of increasing glass packaging disposed into the
environment.
The object of regulating glass packaging disposal is to bring about an optimal
level of waste disposal, which, assuming convexity in the damage and
abatement functions (Perman, Ma and Mcgilvray, 1996: 200-212), is where:
MC r = MC,.
In order to achieve this R.,Rz glass packaging waste must be diverted from the
environment.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
439 SATED NR Vol I (1998) Nr 3
The relevance of all this for government is that, if market generated recycling
only induces a diversion of RoRI waste from the environment, the government
needs, some way or other. to effect an additional diversion of RIR2 waste, in
order to facilitate an optimal level of glass waste disposal.
Instruments of Regulation
There are various ways governments may set about this objective. The most
popular four economic instruments are :
• increasing the imposition of waste disposal charges (a curbside charge on
consumers - based on volume or weight) and fmes for litterirlg,
• levying a product charge (packaging tax) on the sales of producers of
non-refimdable beverage containers, to cover the costs expected to be
incurred in final disposal,
• subsidizing recycling production, and/or
• legislating a deposit-refimd system with respect to all containers (Turner,
Pearce and Bateman, 1993: 259-260).
Each option has problems associated with it (Turner. Pearce and Bateman,
1993: 260-266). For instance, increasing curbside waste disposal charges can
lead to people either disposing of their waste in other ways which negatively
impact on the environment, or adding theirs to someone else's bundle, and it
can be administratively costly. A packaging tax does not directly discourage
irresponsible waste disposal, only indirectly, by raising prices and so reducing
demand. Its main attribute is that it raises revenue to cover the environmental
costs of the waste disposal. Subsidizing recycling systems requires tax payers
to bear part of the consumers disposal cost and so in effect is an income
transfer. The main problem with respect to the deposit refimd system is the cost
benefit studies have not shown it to generate significant net social benefits.
Thecostbenejitapproach
The desirability of the above listed littering abatement measures depends upon
the marginal cost of damage from final disposal being greater than the marginal
cost associated with implementing that particular measure. The issue reduces to
one of whether the marginal cost of damage to the final environment is greater
than the marginal cost of implementing a deposit-refimd system for glass bottle
containers of beverage.
One approach to shedding light on the issue is to conduct a cost benefit analysis
of the proposed legislation at the margin, i.e., using incremental costs. In this
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAmMS NS Vol I (1998) No 3 440
context the cost benefit model advocated by Porter (1978) is popular (Common,
1996:347). The approach is one of using the net present value (NPV) decision
rule - the legislation is only deemed to be desirable if the NPV is positive. For
the case this paper is investigating the NPV would need to be defined as
follows:
T
NPV =L (Bu + BlI + B3l - CII - C21 - C31 )/(1 +rtl
t-I
where,
Bit = the social valuation of the cost saving benefit of the municipality and
other government institutions in collecting and disposing of glass waste,
BlI = the cost saving benefit in reduced inputs required to produce the glass
containers for soft and alcoholic drinks,
B31 = the social valuation of the benefit of fewer glass litter related accidents
and less amenity (or eye-sore) damage,
Cit = the increased cost of inputs used to put soft and alcoholic drinks into
containers and of collecting and administering the expanded refillable
container system,
ClI = the increased cost of greater litter related accidents and amenity (or eye-
sore) damage caused by non-returnable non-glass container substitutes for
glass ones purchased by consumers,
C 31 = the social valuation of the increased time cost to households in returning
empty refillable bottles,
r = the discount rate, and
t = number of years after legislation is implemented; t = L .. T.
There is little doubt that this approach has the potential to yield useful
information, but the information requirements of it are considerable (see Porter,
1978); beyond the terms of reference of this exploratory study. The
estimation of B3l and C31 would require contingent valuations to be carried out
amongst residents and visitors to the area. Many cost benefit studies conducted
overseas on the issue have not yielded significant positive net present values
(Turner, Pearce and Bateman, 1993: 264-265).
An intuitive approach
This paper does not use a cost benefit analysis approach in order to weigh up
the case for bottle deposit legislation, but an intuitive reasoning approach. For
this purpose an expanded version of the model in Figure 1 is adopted - see
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
441 SATEB NR Vol 1 (1998) Nr 3
Figure 2 (overleaf). In it four marginal social cost functions are incorporated:
Me;, Me .. Men! and Mem •
The Me; curve in Figure 2 is as it was for Figure I, but the Mer curve is defined
differently. In the model described in Figure 2, the recycling of non-returnable
bottles is distinguished from that of returnable bottles; the reasons being that the
costs of recycling and the target consumer markets differ. South African
Breweries described the market they provide non-returnable containers for as a
niche one (see section 6 above). Presumably niche markets are catered for by
beverage suppliers because the economies of scale benefit of producing
beverage in standard packaging is outweighed by the increased sales revenue
benefit gained by dividing up the beverage market through different packaging
and advertising. A more complex Mer curve is defined when this division is
taken into account:
Mer = Men!, from Ro to R. (for the returnable share of the glass container
market), and
= Memt from R. to It. (for the non-returnable share of the glass container
market), where,
Men! = the marginal social cost of recycling deposit bearing refillable bottles,
and
Mem = the marginal social cost of recycling non-deposit bearing single fill
bottles (e.g., through the bottle bank system).
The return cost of different types of bottle is assumed to be equal from the
consumer's point of view (implying ell = 0, or that the cost of returning a bottle
to the bottle bank is equal to returning one to a beverage bottle retailer), but not
from the beverage supplier's. As most of the profitable glass packaging
recycling is done in refillable bottles, and the recycling which is being done in
single fill bottles is mainly sustained on moral rather than commercial grounds,
it is deduced that:
The division between the recycling market for returnable and non-returnable
bottles is shown in Figure 2 by the vertical dotted line above R2•
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol I (1998) No 3
Figure 2: A split market model of the case for regulation.
~
Vl
0
U
MC,.!
I
A'
I
o 7
R, R,
MarkCI: non-~clumablc I Markcl: Aelurnable
I
glass conlaiucrs I glass containcrs
MCm
1
35 57
% OF RECYCLABLE GLASS WASTE DISPOSED
INTO THE ENVlRONMENT
442
100
Ro
The case for a mandatory deposit-refund system rests on the benefit of
removing the market division between non-returnable containers and returnable
ones. With the removal of this market division, the efficient level of recycling
increases from R."R2 to R."R). As a result, total costs equal to the shaded area
ABCD are saved; the vertical difference between the MCrd and the lesser of the
MC j and MCm curves. It is a cost which is quite likely to increase steadily in the
future. as citizen opposition mounts to the siting of existing and new tips. so
pushing up tipping costs (Goldstein, 1995: 358).
In principle it is conceivable that this gain through regulation could be offset by
the loss in profits and income gained by the beverage seller and others (like
advertisers) in servicing niche markets. However. as the beverages can be
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
443 SATEB NR Vol I (1998) Nr 3
purchased cheaper from refillable bottles, it appears that much of the profit and
income gained through niche market creation, may be a transfer of consumer
surplus to advertisers and the beverage producing companies, rather than reflect
a net social benefit.
Trends in South African beer and soft drink markets are consistent with this
interpretation. South African Breweries, being overwhelmingly dominant in
the beer industry in South Africa during the 1980s and early 1990s, displayed a
diminishing interest in promoting product differences based on differences in
glass packaging options - more and more of their beer was sold in refillable
bottles. (It is conceivable that this trend may be reversed if new competitors
enter the industry and challenge their market dominance.) In the soft drink
market, by comparison, where competition is more evident, more attention is
focused on differences in packaging.
To sum up: the case for regulating that a deposit-refund system operate in South
Africa for all glass bottle containers of beverage is based on the external cost
saving exceeding the profit gain lost in niche markets. It is a plausible case
because the costs of recycling refillable bottles is less than that of non-refillable
bottles and the profit and income loss in the non-refillable bottle market is
mainly a transfer of consumer surplus to the beverage producers and the
advertisers they commission.
CONCLUSIONS ON REGULATING AGAINST NON-DEPOSIT
BEARING GLASS CONTAINERS OF BEVERAGE
In principle there are many solutions to the problem of glass pollution:
educating consumers to be more responsible, encouraging producers to bear the
burden of the external cost, and placing the burden of this cost on consumers,
by restricting their choice only to glass containers which are deposit bearing.
The latter approach provides consumers with an income incentive to recycle the
glass containers, and so harnesses self-interest as its primary driving force. The
disposal cost of the container is thereby internalized.
It is concluded that it is an efficient option because:
• Private costs would be brought into equivalence with social costs and
most bottle glass waste (over 90%) would never reach final environments.
In principle a litter tax on glass bottle producers can bring about the same
end result (after cleaning up operations) but it would not prevent glass
from reaching final environments.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
SAJEMS NS Vol 1 (1998) No 3 444
• The potential external cost saving is uncontroversial.
Besides being an efficient option, mandatory glass bottle deposit legislation for
alcoholic and soft drinks also may advantage the following:
• local businesses over foreign ones, because imported beverages (like
beer, wine and whisky) would probably need to be packaged locally
instead of abroad in order to fit in with the requirements of the local
deposit-refund system, and
• labour over capital, because a labour intensive production system is
favoured over a capital intensive one in an economy with a huge surplus
of labour.
In reaching this conclusion several potential negative spin-offs of the
proposed regulation were not investigated. The most important of these
spin-offs are increased product prices and substitutions in production and
consumption that aggravate other types of pollution problem, e.g., those
of plastic, tinplated steel and aluminum waste. The prices of beverage in
returnable containers could be raised as a result of reduced physical sales
volumes or the need to recoup forgone profit on the non-returnable
containers. However, it also bears noting in this connection that:
• physical sales volumes of refillable containers of beverage could be
expected to increase, thereby making it possible to derive greater
economies of scale in production, and
• plastic. tinplated steel and aluminum waste problems could (and perhaps
should) be simultaneously addressed, as is commonly done in European
countries where container deposit legislation has been introduced, e.g.,
through the levYing of charges on products using these types of
packaging. The latter approach would also have the merit of being less
distorting with respect to the market for packaging material.
For the above reasons it is recommended that further investigation be conducted
into the merits of introducing mandatory bottle deposit legislation into South
Africa.
ENDNOTE
Funding from the Human Sciences Research Council (HSRC) for this
investigation is gratefully acknowledged, as is the assistance with field work by
Mr. M. Akkerman, Mr. H. Gous, Mrs. J.G. Hosking, Mrs. J. Spearpoint and Mr.
D. Brown, and the comments of the anonymous referees of the HSRC. This
paper is drawn from a research report submitted to the HSRC on 9 February
1998.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.
445 SATEB NR Vol I (1998) Nr 3
REFERENCES
1. BARRY, N. (1994) 'The Market, Liberty and Regulatory State'. Journal
of the Institutefor Economic Affairs, 14 (4), 5-12.
2. BOHM, P. (1981) Deposit-Refund System: Theory and Application to
Environmental, Conservation and Consumer Policy. Baltimore: John
Hopkins.
3. BUSINESS REGULATION REVIEW UNIT (June, 1989) 'Container
Deposit Legislation and the Control of Litter and Waste', Information
Paper No. 14. Report for the Ministry of Industry, Technology and
Commerce, Australia.
4. COMMON, M. (1996) Environmental and Resource Economics: An
Introduction. (Second Edition.) London: Longman
5. GOLDSTEIN, E.S. (1995) Economics and the Environment. New Jersey:
Prentice Hall
6. OLIVER, D.S. (1975) The Use of Glass in Engineering. Oxford: Oxford
University Press.
7. PORTER, R.C. (1978) 'A social cost benefit analysis of mandatory
deposits on beverage containers', Journal of Environmental Economics
and Management, 5, 351-375.
8. RABIE, A. (1976) South African Environmental Legislation. Pretoria:
The Institute of Foreign and Comparative Law, UNISA.
9. SCOTT, W.O. AND COMPANY PTY. LTD. (AUGUST, 1983) 'Study of
the Economic Impact of Beverage Container Deposit Legislation'.
Commissioned by Brewers, Soft Drink Fillers and Container
Manufacturers in Australia.
10. TIETENBERG, T.H. (1992) Environmental and Natural Resource
Economics (Third Edition) New York: Harper Collins.
11. TURNER, R.K., PEARCE, D. AND BATEMAN, I. (1993)
Environmental Economics: An Elementary Introduction. Baltimore: John
Hopkins.
12. VOGLER, J. (1981) Work from Waste. London: Intermediate
Technology Publications Limited.
R
ep
ro
du
ce
d
by
S
ab
in
et
G
at
ew
ay
u
nd
er
li
ce
nc
e
gr
an
te
d
by
th
e
P
ub
lis
he
r
(d
at
ed
2
00
9)
.