A PRICE IS A GUIDE: INTERNALISED MOTIVATIONS AND THE ENGLISH
CARRIER BAG CHARGE
Shaun Larcoma; Luca A. Panzoneb; Timothy Swansonc
a
Department of Land Economy, University of Cambridge;
b
School of Agriculture, Food and Rural Development, Newcastle University, and BENC, Newcastle University;
c
Department of Economics and CIES, Graduate Institute Geneva (IHEID).
Abstract
How does a change in law change behaviour over time? This paper measures the impact of a
legislative change in England that required large retailers to charge for new plastic carrier
bags issued. To measure the behavioural and internalised impacts of the charge, we collected
data prior to the introduction of the charge and two times after its introduction. Using a
difference-in-difference estimator, we found that those who became subject to the charge
used 1.5 new plastic bags less per shop. We also found that introduction of the charge led to
significant changes in internalised attitudes toward plastic bag use. Specifically, we found a
greater acceptance of the government’s role in regulating plastic bag use and an increase in
intrinsic and moral motivation in reducing their use.
Keyword: plastic bag charge; internalisation of law, expressive function of the law,
environmental regulation
JEL codes: Q5; D1; K1; C4.
1. INTRODUCTION
On 5 October 2015 supermarkets and other large retailers in England were required by
law to charge at least 5 pence for single-use plastic carrier bags. The legislative change
followed the introduction of mandatory charges being introduced in Wales in October 2011,
Northern Ireland in April 2013, and Scotland in October 2014.1 The charge was introduced to
reduce the use plastic bags and the litter associated with their improper disposal. In outlining
the rationale for the charge, the UK Government (2015) concluded that single use plastic
carrier bags ‘take longer than other bags to degrade in the environment, can damage wildlife,
and are extremely visible when littered in our towns, parks and the countryside’. It went on to
cite benefits of £6 million per year in savings from litter clean-up costs and £1.3 million per
year in carbon savings. In addition, as retailers are required to donate the funds raised to
‘good causes’, it also included up to £73 million per year to be given to charity.
Prior to the introduction of the mandatory charge, the government relied on voluntary
schemes to try to reduce their use. In 2008 the British Retail Consortium (that represents
larger supermarkets) signed up to a voluntary agreement which saw the distribution of plastic
bags almost half in 2009 compared to their peak in 2006 of 12.1 billion (Commons, 2014,
WRAP, 2015).2 However, since this initial large reduction, plastic bag use began to increase
by approximately 4 per cent per annum to 8.5 billion in 2014 (WRAP, 2015). Indeed, in
explaining its decision to introduce the charge, the UK Government (2015) cited the fact that
their use had increased for a fifth year running.
1
The Republic of Ireland was the first country to require the mandatory charging of plastic carrier bags in 2002,
and since then a number of other jurisdictions across the world have followed suit.
2
Citing data collected from WRAP, the House of Commons report that the reduction was 48 per cent from 2006
to 2009 (compared to the target of 50 per cent). WRAP (2015) report that 12.1 billion plastic carrier bags were
distributed in 2006 and that this fell to 7.2 billion in 2009/10 for the UK. Since then they show an increase each
subsequent year (approximately 4 per cent). The total number of bags distributed in 2014 was 8.5 billion in the
UK (7.6 billion in England).
As noted by the House of Commons Environmental Audit Committee (Commons,
2014, page 7), ‘[d] iscarded plastic bags are an iconic symbol of waste’, and a number of
coalitions were formed in the UK to campaign against their use including the ‘break the bag
habit’ group. Indeed, surveys suggest that support for a mandatory charge has been growing
in England in recent years and that before the charge was introduced, the majority consumers
(62 per cent) were in favour of it.3 Despite a majority of consumers supporting the
introduction of a mandatory charge, some believed that it was an unnecessary intrusion on
consumer sovereignty and there have been reported cases of resistance and non-compliance.
For instance, following the introduction of the charge there were reports of an upsurge in
basket thefts from supermarkets (resulting them being security tagged in some stores) and
there are reports of widespread non-compliance by consumers using self-checkouts that
require consumers to self-report the number of single use carrier bags they use. 4 However,
despite this, early reports by one large retailer suggest that plastic bags issued have fallen by
approximately 80 per cent since the introduction of the charge.5
Given divergent views over plastic bag use, and the staggered introduction of a
mandatory charge across the UK, its introduction to England provides an opportunity to test
the degree that law is internalised by those who are subjected to it.6 In short, to what extent
can a change in the law lead to a change in social values? This empirical investigation also
3
Reported in BBC: Plastic bag charge: Shoppers in England have to pay 5p (5 October 2015).
http://www.bbc.co.uk/news/business-34438030. Accessed January 2016.
4
See on the BBC: Oldham Tesco baskets stolen after bag charge introduced (14 October 2015)
http://www.bbc.co.uk/news/uk-england-manchester-34522034; on The Telegraph: English shoppers steal
£26.7m worth of plastic bags since 5p charge introduced (13 January 2016)
http://www.telegraph.co.uk/news/earth/environment/12097146/English-shoppers-steal-26.7m-worth-of-plasticbags-since-5p-charge-introduced.html; and on The Telegraph: Card Factory dodges government's 5p plastic bag
tax by snipping off handles (28 January 2016)
http://www.telegraph.co.uk/news/earth/environment/12128161/Card-Factory-dodges-governments-5p-plasticbag-tax-by-snipping-off-handles.html
5
See Guardian: Plastic bag usage down 78% since introduction of 5p charge, says Tesco
http://www.theguardian.com/environment/2015/dec/05/plastic-bag-usage-down-78-since-introduction-of-5pcharge-says-tesco. Accessed January 2016.
6
The seminal work of Hart (1997) makes a distinction between the external and internal view of the law, where
compliance can be achieved through both external penalties (even if someone does not accept the law) and
internalisation (where the law is accepted and used to provide guidance for conduct).
provides an opportunity to contribute to the crowding literature, pioneered by Titmuss (1971),
Frey (1992), Gneezy and Rustichini (2000), that highlights how internalised motivations can
be altered through the introduction of legislative change and monetary incentives.7
2. THEORETICAL DISCUSSION
In light of the crowding literature (Frey, 1992, Frey and Stutzer, 2008, Gneezy and
Rustichini, 2000, Gneezy et al., 2011), we can expect that a new law requiring retailers to
charge their customers for new plastic bags to have two effects on their consumption. The
first is the price effect while the second is the internalised (crowding in or crowding out)
effect. That is
xi xip xim


p
p
p
(1)
where 𝑥 represents the number of plastic bags used for consumer i, 𝑥 𝑝 represents the price
effect, and 𝑥 𝑚 = internalised motivation effect, and 𝑝 is the price of the plastic bag. Drawing
on the law of demand, we can expect the price effect to be negative, while the internalised
motivation effect would be negative if crowding in occurred and positive if crowding out
occurred. Therefore, if the charge led to crowding in of internalised motivations, the price
7
While this study is focused on crowding in, there is a much larger body of literature that highlights crowding
out. Crowding out can be generated by control aversion and frustration GNEEZY, U., MEIER, S. & REYBIEL, P. 2011. When and Why Incentives (Don't) Work to Modify Behavior. The Journal of Economic
Perspectives, 25, 191-209, FREY, B. S. & STUTZER, A. 2008. Environmental morale and motivation. In:
LEWIS, A. (ed.) The Cambridge Handbook of Psychology and Economic Behaviour. Cambridge: Cambridge
University Press, BOWLES, S. 2008. Policies Designed for Self-Interested Citizens May Undermine "The
Moral Sentiments": Evidence from Economic Experiments. Science, 320, 1605-1609, GOESCHL, T. &
PERINO, G. 2012. Instrument choice and motivation: Evidence from a climate change experiment.
Environmental and Resource Economics, 52, 195-212.; reduced internal satisfaction or ‘warm glow’ as
individuals no longer feel good about themselves for acting on a voluntary basis (Bowles and Polonia-Reyes
2012, Frey and Stutzer 2008, reduced image motivation from being unable to signal to others that they are a
‘good person’ GNEEZY, U., MEIER, S. & REY-BIEL, P. 2011. When and Why Incentives (Don't) Work to
Modify Behavior. The Journal of Economic Perspectives, 25, 191-209.; release from moral responsibility
FREY, B. S. 1992. Pricing and regulating affect environmental ethics. Environmental and Resource Economics,
2, 399-414.; and frame shifting (e.g. moving the decision from a moral to an economic one, ibid., CARDENAS,
J. C., STRANLUND, J. & WILLIS, C. 2000. Local Environmental Control and Institutional Crowding-Out.
World Development, 28, 1719-1733..
effect would be magnified by internalised motivations. Alternatively, the charge led to
crowding out of internalised motivations, the effect would be ambiguous depending on the
relative size of the effects.
While the literature has focused more on crowding out (Rode et al., 2015, Frey and
Stutzer, 2008), legal scholars such as Cooter (1998) suggest that the law itself can have an
‘expressive effect’ on citizens that can lead to a form of crowding in through a process of
internalisation. Etzioni (2000, page 167) observes that ‘internalisation is a remarkable process
through which imposed obligations (compliance with which must be forced or paid for)
become desires’. Once a law is internalised, some such as Cooter (1998) and McAdams
(1997) suggest that a psychological penalty applies to the act. For example, before the
legislative change, there may have been no psychological penalty associated with using
single use plastic bags when doing one’s grocery shopping, however after, if the norm
concerning limiting their use is internalised, each extra bag used may generate a
physiological cost in the form of guilt or some other negative emotion. Others, such as Hart
(1997, page 242) consider that internalisation or the internal point of view, is manifested by
‘accepting the law as providing guides to their conduct and standards or criticism’ and not
necessarily imposing a psychological penalty.
In terms of how the law can lead to a change in internalised values, various theories
have been put forward. Hart (1997, 89) considers that internalisation involves a shift in
viewpoint; that sees one move from ‘observer’ to a ‘member’ and then the member uses the
law as a guide for his or her conduct and in determining what is good (or right) behaviour.
Hart’s analysis implies that if one does internalise the law, then a legal change will induce
changes in attitudes toward the behaviour in question as well as the regulatory space occupied
by the legislator. For instance, citizens may internalise both the norm of minimising plastic
bag use and the state’s authority in aiming to minimise plastic bag use through imposing a
mandatory charge. Somewhat differently, Cooter (1998, page 608) considers that ‘coercive
state sanctions can induce people to internalise norms by creating opportunities for Pareto
self-improvement’. He suggests that a Pareto self-improvement is generated when
commitment to the new norm produces an advantage relative to an individual’s original
preferences, which can be generated through a new social equilibrium that favours those who
adhere to the new norm. In this sense, a legislative change signals the benefits of adopting a
new norm as well as generating a new social equilibrium. While strictly not a form of
internalisation, but closely related to Cooter’s analysis, another rationale for crowding in may
occur in relation to a charge being introduced is due to increased returns from prosocial/environmental behaviour generated by correcting for a co-ordination (or free-rider)
problem. For example, as noted by Rode et al. (2015, page 271) ‘it’s easier for
environmentally minded car drivers not to speed on the highway when a general “speed”
limit and fine prevents other drivers from overtaking them’.
Based on the discussion above, we can expect to observe both a price effect and
internalisation effect generated by the introduction of the mandatory carrier bag charge. In
terms of the price effect, we can expect this to have a negative effect on consumption. In
terms of internalised motivations, if law is internalised, the legislative change should see an
increase in intrinsic motivation concerning minimising plastic carrier bag use. Under Hart’s
analysis, this internalisation should be generated by the mere fact that law is internalised and
used as a guide for conduct and what is right. Indeed, in this sense internalisation of the law
can take two aspects: one focused on the rightful place of government involvement in
consumer/business decisions (i.e. the Government is right to interfere in the market) and by
modifying their attitudes toward use of plastic bags. If internalisation is generated under
Cooter’s analysis, we should see beliefs around increased returns from plastic bag
minimisation increase following the legislative change.
3. ECONOMETRIC MODEL
The objective of this exercise is to determine the impact of a legislative change that
mandates a charge on new plastic carrier bags, both in terms of behaviour and internalised
motivations. To this extent, imagine a consumer i of characteristics X i in week t, where t =
0 before the policy is introduced, and t = 1 indicates a period after the implementation of the
charge. The population is further divided into two groups s. Some consumers (s = 0) already
pay a charge on plastic bags on environmental grounds at t = 0, and they are not affected by
the policy (in experimental terms, they are the “control group”). Others (s = 1) are not
exposed to the policy at time t = 0, and the policy influences their behaviour (in experimental
terms, they are the “treated group”). Total plastic bag consumption of individual i in group s
at time t is then be defined as 𝑌𝑖𝑠𝑡 . The total impact of the policy on the target consumers can
then be written as the difference in consumption of individuals previously unexposed to the
charge compared to the same difference in an identical individual who was previously
exposed to the charge (Imbens and Wooldridge, 2009), as
ˆ  Yi (s  1, t  1)  Yi (s  1, t  0)  Yi (s  0, t  1)  Yi (s  0, t  0)
(2)
Equation (2) represents a difference-in-difference (DID) estimator. This estimator can
be implemented in practice by estimating the regression
𝑌𝑖𝑠𝑡 = 𝐴𝑠 + 𝐵𝑡 + 𝛽𝐼𝑠𝑡 + 𝛿𝑋𝑖𝑠𝑡 + 𝜖𝑖𝑠𝑡
(3)
where A refer to group specific fixed effects (e.g. different habits of the control compared to
the initially unexposed group), B refer to time-specific fixed effects (e.g. the different
coverage of the policy in the media in the different periods), while X refer to all the other
covariates influencing the decision of the consumer (e.g. demographics). In equation (3), I
refers to the interaction term between treated participants (those initially not exposed to the
charge) and the legal change dummy. Then, β capture the net impact of the legal change
(henceforth ‘policy’), measuring the change in consumption of people exposed to the policy
for the first time relative to the control group. Importantly, the presence of different time
periods causes problem of serial autocorrelation of the residuals 𝜖𝑖𝑠𝑡 (Bertrand et al., 2004);
this problem can be significantly mitigated by clustering residuals by period (Wooldridge,
2003).
4. DATA
4.1. SURVEY DESIGN
To measure the impact of the plastic bag charge on English consumers, data on the
consumption of plastic bags and attitudes was collected through three subsequent crosssectional surveys of the UK population. The requirement to charge for plastic bags came into
force in England on October 5th, 2015. To assess consumption before the introduction of the
charge, a first survey took place on the 29th-30th of September 2015. A second survey
collected data one week after the introduction of the policy, on the 13th-14th of October 2015.
A final round was then set seven weeks after the introduction of the carrier bag charge (24th25th November, 2015), to capture the mid-term impact of the policy on behaviour. This
survey, identical across periods, collected information on the consumption of plastic bags,
plus personal attitudes towards plastic bag use and its regulation. To obtain nationally
representative samples, the survey used a quota-sampling strategy that set quotas for age,
gender, and regional distribution of the population to match the actual UK population. A
median test shows that respondents across the three periods also do not differ significantly in
terms of income (p=0.474), and education (p=0.716). Finally, all participants were able to
complete the survey within 5 minutes.
The need to collect a survey to understand consumption is driven by the lack of
available information on plastic bag consumption. Before the introduction of the charge,
plastic bags were free to take and they were not scanned at the till (they had no bar code).
This situation changed once the charge was introduced, because consumers had to pay for the
bag and this charge appeared on the bill. To ensure the availability of the same consumption
data before and after the policy was introduced, the survey used a standard recall question
(Browning et al., 2003): consumers were asked to report the number of carrier bags they
needed in their last grocery shopping trip when they used some. Self-reported measures of
behaviour are known to carry a downward bias (Schwarz, 1999, Browning et al., 2003),
particularly when consumption is considered socially undesirable, e.g. alcohol (Feunekes et
al., 1999). However, consumption from recall can be accurate if the questionnaire is designed
correctly: recalling a specific event, which has occurred recently, and having no time
restriction to answer the question has been shown to give more precise measures of behaviour
(Schwarz and Oyserman, 2001). As a result, respondents were asked to base their assessment
in the last grocery shopping trip and were given a scale to facilitate the task. The same recall
approach was used to assess supermarket brand preferences, donations to charity, availability
or purchase of a reusable/long-life carrier bag, and the purchase of plastic bin liners.
To test the process of internalisation of the new law, the survey collected information
on two aspects: personal attitudes toward plastic bag use; and the regulatory space of
government. Unless otherwise stated, these variables were collected on scales going from
zero to 100. In particular, the survey measured: intrinsic motivation, asking the agreement to
the statement “Minimising the number of plastic carrier bags when I shop for groceries is
important to me, regardless of any benefit or inconvenience that may result”; warm glow,
asking “I feel good when I don’t use new plastic carrier bags when I shop for groceries
because it helps the environment”; and perceived morality of the task, based on the
statement “Shoppers have a moral obligation to minimise the use of plastic carrier bags”. To
understand the perceived role of the regulatory space, the survey measured: internalisation,
using a binary variable that asked “Do you think it is right for the government to make
retailers charge customers for plastic carrier bags?” and the perceived role of the
government, asking the agreement to the statement “The government should not interfere by
requiring retailers to charge for plastic carrier bags”.
A final set of questions measured: the perceived policy effectiveness, asking the
agreement to the statement “Charging for carrier bags and giving the profits to a good cause
(e.g. donating to a charity) will ensure that plastic carrier bag use is no longer a problem for
our society”; the perception of the problem based on the statement “Plastic carrier bags are
currently overused”; and peer comparison, asking the agreement to the statement “Other
customers will continue using plastic carrier bags even if I stop using them”. In an effort to
focus our results on research aims and due to space limitations, the analysis is focused on
stated behavioural responses and internalised attitudes toward plastic bag use.
4.2. IDENTIFICATION STRATEGY
As defined in section 3, the impact of the plastic bag charge is identified using a differencein-difference estimator. This type of estimator requires variability in the sample along two
dimensions: firstly, the sample should contain information of behaviour before and after the
policy is introduced; secondly, in both periods the sample should include a number of
individuals who are not affected by the policy when this is implemented, which represent a
control group against which changes are compared (see also Bertrand et al., 2004 for more
details on the method). The present dataset contains information on plastic bag consumption
before as well as after the charge was introduced. The treatment-control dimension of the
estimator is captured by policy variation within UK supermarkets. Firstly, the market
presented institutional differences: specifically, Wales introduced a plastic bag charge on the
1st of October 2011, followed by Northern Ireland on the 1st of April 2013, and from Scotland
on the 20th of October 2014. At the same time, one retailer chain introduced the charge
voluntarily: Marks & Spencer started charging 5p for plastic bags in the UK using an explicit
environmental rationale since February 2008 and donating the resulting profits to charity8.
As a result, UK supermarkets represent a setting that lends itself to a quasiexperimental design: the plastic bags charge mattered only for English consumers who do not
shop in Marks & Spencer. The remainder of the sample is the control group, which captures
those consumers who were already paying for plastic bags on environmental grounds. Table 1
observes that around one third of consumers had shopped in Marks & Spencer in the four
weeks leading to the survey, and the sample shows no major shift in the choice of retailer
brand, with consumers on average visiting about 4 different store brands a month. Notably,
the control group and the treatment group might differ in terms of unobservable preferences
for the environment even before the most recent plastic bag charge was introduced. As a
result, a different customer base should favour the incorrect rejection of a true null hypothesis
(a type I error). Similarly, the incorrect inclusion of individuals who are not exposed to the
policy in the control group would result in significant heterogeneity in response to policy,
leading to the failure to reject a false null hypothesis (a type II error).
Table 1: General shopping habits of the sample
Store
Survey 1
Survey 2
Survey 3
Asda
47.00%
47.56%
46.35%
Aldi
35.78%
34.33%
33.85%
Coop
27.00%
25.33%
25.66%
8
See http://corporate.marksandspencer.com/blog/stories/carrierbagcharge, and
http://www.thisismoney.co.uk/money/bills/article-3254567/Clothes-shops-charge-5p-carrier-bag-Monday.html
Iceland
21.44%
20.11%
22.12%
M&S
33.00%
31.67%
32.63%
Morrisons
31.56%
34.22%
34.96%
Lidl
31.22%
28.78%
29.42%
Sainsbury
50.00%
50.00%
55.20%
Tesco
64.00%
65.22%
64.49%
Waitrose
20.11%
18.78%
17.92%
Small local retailer
17.11%
16.22%
17.15%
Other stores
2.89%
3.56%
3.76%
Average number of stores visited
3.81
3.75
3.83
Observations
900
900
904
5. RESULTS
This section estimates the impact of the introduction of the plastic bag charge on consumer
attitudes and behaviour. The section starts by making an assessment of the overall change in
variables across the survey, and the change in consumption of plastic bags over time.
Subsequently, the section uses the DID estimator to quantify the motivational and
behavioural impact of this policy.
5.1. CONSUMER ATTITUDES TOWARDS THE PLASTIC BAG CHARGE
In this section, we first summarise the characteristics of the samples. Table 2 indicates that
the response to several attitudinal and behavioural variables changed over time. Specifically,
one week after the carrier bag charge came into force (i.e. in the second survey), support of
the policy increased. In some cases, this support seems to fade after a few weeks, a point
which is formally tested in the difference-in-difference analyses below. Table 3 also shows
that before the policy was implemented, behaviours did not vary enormously across UK
country and between treated and controls: total consumption of bags, and the purchase of
“bags for life” and bin liners did not differ significantly, while consumers paying for bags use
fewer new bags, and are more likely to have reusable bags accessible to consumers. As a
result, the samples did not differ in their plastic bags needs, but in their ability to reuse bags.
However, all attitudinal measures differed significantly across UK country (table 3): English
consumers appeared less motivated to reuse bags, considered plastic bags use less
problematic, and were more likely to dislike governmental intervention than countries where
the policy was already in place. The same attitudinal base did not vary as much between
English controls and treated groups: controls were only more likely to consider that
government intervention was right, they felt better after reusing plastic bags, and were more
sensitive to the problem of plastic bags overuse.
Table 2: Mean attitudinal variables in the sample
Variable
Behaviours
Attitudes
Survey 1
Survey 2
Survey 3
Bag used – All
4.4878
4.3689
4.1836
Bag used – New
2.5467
1.9067
1.4447
Reusable bag – carried
0.6233
0.7178
0.7478
Reusable bag – purchased
0.2344
0.3122
0.2777
Bin liners purchased
0.2056
0.2400
0.2290
Government is right to make retailers charge for plastic
bags
Fair charge (£)
0.664
0.722
0.698
5.049
7.083
7.149
Minimising plastic bags is important
64.742
66.888
65.349
I feel good when I don’t use new plastic bags
67.313
69.571
68.801
Charging for bags and giving the profits to a good cause
removes the problem
The government should not interfere by requiring
retailers to
charge for plastic carrier bags
Plastic carrier bags are currently overused
62.151
61.916
61.996
46.107
40.823
43.016
71.328
73.309
69.166
Other customers will continue using plastic bags even if I
stop
Moral obligation to reuse bags
70.872
67.503
66.982
68.293
70.072
68.478
900
900
904
Observations
Table 3: Comparison of measured variables across groups and countries at week 1
Group (England only)
Control
Treated
a
Country
England
N. Ireland
Scotland
Wales
chi2 (3)b
Behaviours
Attitudes
Total bags used
4.2875
4.7237*
4.5849
4.0000
4.1286
3.7755
6.403*
New bags used
Reusable bag – carried
2.4583
2.9533*
2.7958
0.9259
1.1714
1.5714
48.387***
0.6958
0.5447***
0.5928
0.8889
0.8143
0.6735
22.466***
Reusable bag – bought
0.2208
0.2471
0.2387
0.1852
0.1714
0.2857
2.705
Bin liners
0.2125
0.1848
0.1936
0.2963
0.3000
0.2041
5.835
Government is right
0.6917
0.6070**
0.6340
0.7778
0.8571
0.7959
20.134***
Fair charge
5.0667
4.5136
4.6897
7.2593
8.2857
4.7347
43.553***
Intrinsic motivation
65.8875
62.2899
63.4350
75.2963
72.1429
68.4694
11.198**
Warm glow
69.3208
63.9591**
65.6658
81.3333
77.6286
70.2041
23.891***
Good cause solves problem
61.6583
59.5798
60.2414
82.2222
70.2571
68.8980
26.672***
No gov't interference
47.5792
49.6965
49.0226
29.6296
30.0143
33.3061
36.087***
Bags are overused
74.8208
69.0992***
70.9204
80.2593
75.3714
66.8980
7.257*
Peer comparison
71.6083
72.6031
72.2865
63.4074
62.6000
65.0408
14.443***
Moral obligation
68.8750
65.6673
66.6883
78.2593
78.7429
72.5714
20.092***
240
514
754
27
70
49
Observation
a: Significance refers to an independent two-sample t-test with unequal sample sizes and unequal variances. b:
significance based on a Krusal-Wallis test.
5.2. CONSUMPTION OF PLASTIC BAGS
A first question related to the policy discussed in this article is the impact of the policy on
plastic bag consumption. The target of the policy was to reduce new plastic bags used, and
increase the reuse of existing bags. Table 2 shows that total plastic bag consumption
decreased very mildly in the weeks following the introduction of the policy, going from 4.5
to 4.2 units per person in the UK. Conversely, consumption of new plastic bags dropped from
around 2.5 units per person to less than 1.5 bags, corresponding to a 43% decline. Figure 1
shows the distribution of the consumption of both new and all plastic bags in the three survey
periods. A Kolmogorov-Smirnov test (table 4) shows that the distribution of total plastic bags
used did not change significantly over time, while the distribution of new bags decreased.
Figure 2 indicates that individuals previously unexposed to the policy showed both a decline
in total as well as new plastic bags (p values for trend equal to 0.022 and 0.000, respectively);
conversely, the introduction of the policy in England did not change behaviour in consumers
already exposed to the charge (p-value for trend = 0.901), but also reduced significantly their
uptake of new plastic bags (p-value for trend = 0.000).
Comparing the impact of the policy on those consumers initially exposed and
unexposed to the plastic bag charge, figure 3 confirms that the consumption of plastic bag
(new and used) went down slightly both in consumers who were initially exposed to the
plastic bag charge and those who were not. An ANOVA test indicates that the interaction
between treatment and policy is significant in period three only for total bag consumption (pvalue = 0.0279); while for new bags the interaction is significant both in period 2 (p-value =
0.0735) and period 3 (p-value = 0.0001). As a result, the introduction of the mandatory
charge was successful in reducing the consumption of new plastic bag use very rapidly, but
changes in the overall consumption of plastic bags declined more slowly, reaching a
significant decline somewhere between two and seven weeks after the policy was introduced.
Finally, figure 3 graphically shows convergence between the consumption of new plastic
bags in both consumers initially exposed to the charge (control) and those initially unexposed
(treated) after 7 weeks.
Table 4: Kolmogorov-Smirnov test
Week 1 vs Week 2
Week 1 vs Week 3
Increase in plastic bags
Decrease in plastic bags
Increase in plastic bags
Decrease in plastic bags
New plastic bags
D Statistic
0.0000
-0.1011***
0.0000
-0.2328***
Figure 1: Distribution of plastic bags used, by survey period
a) Period 1 (pre-policy)
Total plastic bags
D Statistic
0.0244
-0.0278
0.0000
-0.0403
1.2
1
.8
.6
0
.2
.4
Density
0
5
10
All bags
15
New bags
.6
.4
.2
0
Density
.8
1
1.2
b) Week 2 (post-policy)
0
5
10
All bags
c) Week 3 (post-policy)
New bags
15
1.2
1
.8
Density
.6
.4
.2
0
0
5
10
All bags
New bags
Figure 2: Trend in consumption of plastic bags over time
15
Figure 3: Average number of plastic bags used by survey period and treatment
0
1
2
3
a) New plastic bags
Survey 1
Survey 2
Control
Survey 3
Treated
6. DIFFERENCE-IN-DIFFERENCE ANALYSIS
The previous analysis concentrated on descriptive statistics and univariate test of differences
in motivational and behavioural variables in the sample. This section uses the difference-indifference estimator described in section 3 to determine the net impact of the English plastic
bag charge. In the regressions described in this section, demographic characteristics enter as
follows: age is used as a linear variable, gender refer to a dummy equal to one for males,
while income and education are used as a set of dummies. Finally, attitudinal variables,
which were measured using a continuous variable (going from 0 to 100), are used in linear
form. In the analyses in this section, the control group includes individuals who were initially
exposed to a plastic bags charge: it groups consumers who have shopped at least once in
M&S in the four weeks previous to the survey; and respondents from Wales, Scotland, and
Northern Ireland. The remainder of the sample represents the “treated” group.
6.1. CHANGE IN THE CONSUMPTION OF PLASTIC BAGS
The first part of the analysis refers to the impact of policy on the consumption of plastic bag.
In fact, the primary objective of the policy was to “reduce the use of single-use plastic carrier
bags, and the litter they can cause, by encouraging people to reuse bags 9”. Consequently, the
introduction on a charge on plastic bags would be expected to motivate a reduction in new
plastic bags taken from stores in favour of the reuse of older bags available to the consumer.
As a result, this section measures the impact of the plastic bag charge on consumer
behaviour, observing both new and total plastic bag consumption. Table 5 indicate that the
policy reduced consumption by around 0.5 bags per person per trip after seven weeks of
policy, down from a smaller drop (around 0.09 bags) after one week. On the other hand, and
as expected, the policy had a much stronger impact on the use of new plastic bags, which
declined by over 1.5 plastic bags per person per trip after seven weeks, a significant drop
from the 0.3 reduction in consumption one week after the policy was introduced.10
Importantly, figure 5 shows that intrinsic motivations and warm glow are among the
drivers of this reduction in the use of new plastic bags, but have no influence on total plastic
bags used. On the other hand, support for governmental interference, peer comparison, and
attitudes towards plastic bags overuse favour a reduction in both new and total plastic bags
used. Surprisingly, feeling the moral obligation to recycle plastic bags did not influence
behaviour. Finally, optimistic expectations over the effectiveness of the policy increased
plastic bags use: the belief that donating the charge to a good cause will remove the problem
appears to stimulate the consumption of both total and new plastic bags.
9
https://www.gov.uk/government/publications/single-use-plastic-carrier-bags-why-were-introducing-thecharge/carrier-bags-why-theres-a-5p-charge
10
In relative terms, this drop corresponds to a 10% decline in the share of new bags in the last survey period.
Notably, the reduction in total bags used might be driven by the larger size of new plastic bags currently
available in the market, as well as to the larger storage capacity of reusable bags.
Table 5: Impact of the plastic bag charge on plastic bag consumption
All bags used
New bags used
Coefficient
S.E.
Coefficient
S.E.
Intercept
7.3748***
0.1638
8.7087***
0.4778
Control group fixed effects
0.3595***
0.0597
0.8808***
0.0892
0.0311
0.0420
-0.8528***
0.1067
-0.0873***
-0.0916***
0.0168
0.0326
-0.3293***
-1.6856***
0.0387
0.1043
Age
-0.5205***
-0.0350***
0.0517
0.0032
-1.6163***
-0.1004***
0.0860
0.0112
Male
-0.4923*
0.2533
0.4320
0.4633
Intrinsic motivation
-0.0086
0.0086
-0.0234***
0.0050
Warm-glow
0.0004
0.0046
-0.0129***
0.0037
Optimism
0.0094***
0.0033
0.0211***
0.0009
Government interference
0.0105***
0.0030
0.0251***
0.0041
Bags are overused
-0.0083**
0.0040
-0.0128*
0.0073
Peer comparison
-0.0044***
0.0010
-0.0095***
0.0017
Moral obligation
0.0007
0.0021
-0.0028
0.0017
Income dummies
Yes
Yes
Education dummies
Yes
Yes
Survey period 2
Policy impact – period 2
Survey period 3
Policy impact – period 3
3.2432***
Sigma
Observations
Pseudo R
2
Log-likelihood
0.0882
4.6010***
2704
2704
0.0180
0.0635
-6391.56
-4729.60
0.2075
Note: residuals are clustered by period
6.2. INTERNALISATION OF THE POLICY
The previous section observed a significant impact of the plastic bag charge on consumer
behaviour, particularly on the uptake of new plastic bags. This section now considers the
impact of the introduction of the carrier bag charge on consumer attitudes. The section
focuses on two main aspects of the legislative change: the internalisation of the role of the
state, in terms of whether governmental actions are the right mean to achieve the objective;
and the perception of the role of the consumer in dealing with this environmental problem.
6.2.1. THE REGULATORY SPACE
Table 6 shows that the overall decline in the consumption of new plastic bags is accompanied
by changes in the way consumers view the role of government in tackling climate change. In
particular, consumers increasingly believe that government is right in making retailers charge
customers for plastic carrier bags. The marginal impact of the policy, reported in table 7,
appears to be around 2% one week after the introduction of the policy, reaching an increase
of around 8% after seven weeks. Similarly, the attitude against the interference of the
government in charging for plastic carrier bags decreased by 1.1 units after one week, and
dropped by 7 units after seven weeks. As a result, consumers started viewing the role of
government in dealing with excessive plastic bags use positively, leading also to increased
willingness to contribute to this regulation (figure 4).
Table 6: Impact of the plastic bag charge on the internalisation of the policy
Gov’t interference
Internalisation
Coefficient
S.E.
Coefficient
S.E.
-0.0845
0.1524
53.9861***
1.2431
Control group fixed effects
-0.3431***
0.0125
7.7380***
0.1193
Survey period 2
0.1117***
0.0101
-4.3670***
0.3088
Policy impact – period 2
0.0755***
-0.0462***
0.0148
0.0086
-1.1706*
0.8328*
0.3469
0.2130
Age
0.2403***
0.0077***
0.0156
0.0013
-6.9806***
-0.1170*
0.2202
0.0303
Male
-0.0858
0.0728
0.1170
0.4616
Intercept
Survey period 3
Policy impact – period 3
Income dummies
Yes
Yes
Education dummies
Yes
Yes
2,704
2,704
Observations
Pseudo R
2
Log-likelihood
R
2
Note: residuals are clustered by period.
0.0286
-1615.68
-13301.67
0.0270
Table 7: Marginal effects from probit equations
Internalisation
Coefficient
S.E.
Policy impact – period 2
0.0251***
.0033
Policy impact – period 3
0.0796***
.0088
8
Figure 4: Fair charge for a plastic bag over time
7.15
2
3
6
7.08
0
2
4
5.05
1
Note: figures refer to averages. Differences between period 1 and the two post-policy periods are significant at
5% (period 2) and 6% (period 3) using a one-tailed t-test. Kruskal-Wallis rank test: chi2 = 14.98 (p = 0.0006).
6.2.2. ATTITUDES TOWARDS PLASTIC BAGS USE
The second set of results illustrates the impact of the plastic bags charge on internalisation, in
terms of the acceptance of the legislative change in reducing plastic bags consumption. Table
8 shows that intrinsic motivation (the belief that plastic bags reuse is correct, independently
on costs and benefits of doing it) increased after the policy was introduced, and remained
significant after seven weeks. Similarly, warm glow, intended as feeling good when not using
new plastic carrier bags when shopping on environmental grounds, increased significantly
after one week, but its effect halved after seven weeks. Finally, the legislative change
successfully increased the perception among consumers that plastic bags reduction is a moral
obligation, with a sharp increase after one week that remained significant over time. In
summary, these results suggest that the policy strongly communicated to consumers the a
plastic bags problem, but the strong attitudinal response seems to reduce (mildly) in
magnitude as time progresses (the difference is however significant only for warm-glow),
possibly reaching a stabilisation point as consumers learn about their preferences.
Table 8: Impact of the plastic bag charge on consumer attitudes towards plastic bags saving
behaviours
Model
Intrinsic motivation
Warm-glow
Moral obligation
OLS
OLS
OLS
Coefficient
S.E.
Coefficient
S.E.
Coefficient
S.E.
Intercept
49.1513***
1.9175
60.5465***
2.3729
58.8379***
2.3254
Control group fixed effects
-4.7418***
0.1815
-7.3763***
0.1122
-5.3528***
0.1724
0.8262
0.2839
-1.6333***
0.0770
-0.7095**
0.1097
1.9062**
-0.2530
0.3787
0.0955
6.3664***
-0.6734***
0.1824
0.0663
3.8353***
-1.5289**
0.1961
0.1651
Age
1.5522**
0.2734**
0.1564
0.0297
3.6974***
0.1898**
0.0094
0.0270
2.8672***
0.2141**
0.2396
0.0244
Male
-4.0102*
1.0306
-5.5132**
1.2360
-3.5441**
0.5721
Survey period 2
Policy impact – period 2
Survey period 3
Policy impact – period 3
Income dummies
Yes
Yes
Yes
Education dummies
Yes
Yes
Yes
Observations
2704
2704
2,704
Log-likelihood
-12810.40
-12765.38
-12617.61
0.0367
0.0341
0.0325
R2
Note: residuals are clustered by period.
6.2.3. SOLVING THE COORDINATION PROBLEM
A final set of result explores the impact of the plastic bag charge on consumer views over
internalisation. In particular, the policy focuses on two aspects: the view that plastic bag
overconsumption is a problem; and the expected behaviour of plastic bags on other
consumers. Results (table 9) indicate that policy contributes to addressing a coordination
problem. Firstly, the introduction of the plastic bag charge increases the relevance of the
problem: when introduced, the policy immediately raises the perception of plastic bags being
overused. The increased salience of the problem remained significant even after 7 weeks after
the introduction, when consumption of plastic bags had started to significantly decrease, as
from table 511. At the same time, the introduction of the charge reduced reliance on selfcomparison in long-run. Specifically, consumers showed an increased scepticism in the
ability of other consumers to reduce their plastic bags overconsumption, but the value
dropped significantly after 7 weeks. As a result, the policy made both consumers more aware
of a plastic bag problem and more optimistic on the collective ability to deal with it.
Table 9: Impact of the plastic bag charge on consumer attitudes towards plastic bags use
Bags are overused
Peer comparison
Coefficient
S.E.
Coefficient
S.E.
Intercept
58.4854***
2.3613
70.5729***
2.8524
Control group fixed effects
-4.5331***
0.1406
3.5049**
0.4009
Survey period 2
-0.8274**
0.1633
-4.1348***
0.2175
Policy impact - period 2
4.5314***
-4.6509***
0.1775
0.0755
1.2716**
-1.5214**
0.2276
0.1958
Age
4.4576***
0.1758***
0.0924
0.0150
-4.2571***
0.0034
0.2126
0.0361
Male
-2.7566
1.1233
-2.2893*
0.6865
Survey period 3
Policy impact - period 3
Income dummies
Yes
Yes
Education dummies
Yes
Yes
11
See also http://www.telegraph.co.uk/news/earth/environment/12034574/Plastic-bag-5p-charge-cuts-usage-byalmost-80-per-cent-first-data-shows.html.
Observations
2704
2704
Log-likelihood
-12598.19
-12169.64
0.0268
0.0278
R2
7. ROBUSTNESS TESTS
The difference-in-difference results presented in the previous section estimated the impact of
the policy on a group of consumers unexposed to the charge relative to a reference group who
were already paying the charge. The choice of reference group might influence the results
observed in this analysis. This section presents a series of robustness checks to test the
validity of the results presented here. In particular, the analyses above are repeated for three
alternative control groups. The first alternative is to observe change only in England (where
the new policy was introduced), removing Scotland, Northern Ireland, and Wales; in this
case, the control group refers to shoppers in M&S, who were already being charged for
plastic bags for environmental reasons. The second option is to remove individuals shopping
in M&S, comparing inter-country differences; using this control group would remove the
potential self-selection of environmentally motivated shoppers who may prefer M&S because
of their relatively high environmental standards. Finally, a last control group is identical to
the control of the main analysis, but also includes shoppers in Lidl and Aldi, as they were
also paying for plastic bags, although not on environmental grounds, and profits were not
donated to a good cause.
Table 10 presents the policy coefficients using the different control groups. The impact
of the policy does not seem to change qualitatively, albeit estimates do differe in quantitative
terms. Specifically, the consumption of plastic bags decreased independently on the control
group considered. However, when considering England only, it seems that in the short-term
the policy resulted in an increase in new plastic bags used, which then decreased significantly
after seven weeks. Internalisation and support for the governmental encroaching in the
regulatory space increased in all groups, although the change was not significant when
considering England only, and actually decreased when excluding M&S customers, but the
long-run effects coincide in all three groups. Intrinsic motivation instead increased only in the
sample excluding M&S, suggesting that cross-country differences are more marked than
within-country variations in preferences. A similar argument appears for moral obligation,
which is consistently strongly significant when M&S customers are excluded. Finally, warmglow does not appear to vary in substance across sample.
Table 10: Robustness checks for the difference-in-difference estimator
Policy impact
Previous results
Only England
No M&S
Incl. Lidl & Aldi
Period
Coefficient
S.E.
Coefficient
S.E.
Coefficient
S.E.
Coefficient
S.E.
2
-0.3293***
0.0387
0.2622***
0.0451
-1.4151***
0.0396
-0.3929***
0.0270
3
-1.6163***
0.0860
-1.0684***
0.0765
-2.0119***
0.1106
-1.4941***
0.0907
2
0.0755***
0.0148
0.0122
0.013567
-0.1328***
0.0119
0.0657***
0.0112
3
0.2403***
0.0156
0.1553***
0.0230574
0.3269***
0.0189
0.2528***
0.0154
Intrinsic
2
1.9062**
0.3787
0.7415
0.4195
1.7529**
0.3726
0.5115
0.3447
motivation
3
1.5522**
0.1564
0.2055
0.2981
4.3673***
0.2770
0.6382
0.3282
Warm glow
2
6.3664***
0.1824
5.6372***
0.1352
5.4763***
0.3369
3.9725***
0.0973
3
3.6974***
0.0094
2.0596***
0.1750
4.7893***
0.1583
1.9550**
0.2520
Government
2
-1.1706*
0.3469
3.0089**
0.4016
-5.4877***
0.4813
-0.5820
0.3266
Interference
3
-6.9806***
0.2202
-2.8361**
0.4382
-13.2585***
0.2670
-5.0458***
0.4317
Moral obligation
2
3.8353***
0.1961
2.6228**
0.2694
3.1756***
0.2942
1.8079**
0.2212
3
2.8672***
0.2396
0.4747
0.3422
5.2406***
0.2976
1.2486*
0.4132
2
1.2716**
0.1775
4.1549***
0.1693
-1.3646***
0.0765
1.3208
0.1498
3
-4.2571***
0.0924
-1.3333**
0.2716
-6.4315***
0.3483
-4.9487***
0.2332
2
4.5314***
0.2276
5.6239***
0.3435
1.3236**
0.2161
1.9893**
0.2551
3
4.4576***
0.2126
3.7291***
0.2298
4.2071***
0.4180
3.4251***
0.2349
New bags
Internalisation
Peer-comparison
Bags are overused
Observations
2,704
2,265
1,827
2,704
8. DISCUSSION
This article aims to measure the impact of the introduction of a mandatory plastic bag charge
on English consumers. The policy was introduced England in October 2015, and an identical
policy was already in place in all remaining UK countries. Similarly, one UK retailer was
voluntarily charging for plastic bags on environmental grounds at the time the policy was
introduced. This setting provides an ideal context to test the net impact of a policy by
comparing the response of consumers who were already exposed to the charge when the
policy was implemented against those who were not, using a difference-in-difference
estimator. Results indicate that the policy had a mild impact on total plastic bags use, but a
much stronger effect on the use of new plastic bags.12 Indeed, new plastic bag consumption
fell dramatically and broadly in line with other countries that implemented similar charges,
both across the UK and further abroad. In this sense the legislative reform has achieved its
aim.
More importantly, this study sheds some light on how the legislative change was
internalised by consumers. The plastic bags charge appealed to the need for a reduction in the
consumption of new bags. The changes in consumption observed in the results of this article
appear to accompany changes in the way the policy is interpreted and understood. In
particular, the policy had very strong impact on the attitudes of consumers. Results support
evidence of an internalisation process, whereby consumers took on the rationale for the
legislative change once it was introduced, and supported the legislator in increasing the
regulatory space within which it operates. Such a finding provides strong empirical support
for Hart's (1997) conception of internalisation and the process internalisation. A significant
12
This finding somewhat conflicts with earlier evidence, which showed a spillover effect that decreased total
consumption of plastic bags once the charge was introduced in M&S (Disney et al., 2013): while this finding is
in line with the results presented here, the spillover effect is fairly small, and might be smaller if it accounts for
the larger size of new bags after the policy, because by default they would reduce the need for bags.
portion of English consumers to use the law as a guide for conduct and what is right.
Interestingly, as the questions relating to solving co-ordination problems associated with
plastic bag use in some cases moved in the opposite direction in the short term (i.e. one week
after the policy was introduced), but over time consumers seemed to perceive improved
incentives in addressing the coordination problem. These results suggest that the
internalisation process was largely generated by a change in the law itself.
Another important result was the temporal aspect of the internalisation process. The
results indicate that consumers internalised the legislative change immediately, i.e. one week
after its introduction. However, while the internalisation of the regulatory space increased
over time, the motivational and moral views of plastic bags reuse decreased after seven
weeks compared to the first week, albeit remaining strongly significant. In other words, the
introduction of legislative change led to increasing support for the role of the government
over time; conversely, attitudes towards plastic bag use showed an overly optimistic pattern,
with a big jump when the policy is introduced that settled back down as time passed. These
results highlight the complexity of the internalisation process and how some aspects of
internalisation (and crowding more generally) seem to be short lived. This is consistent with
the findings of Goeschl and Perino (2012) who found the crowding out effects of an
emissions tax (in a laboratory setting) to be only temporary.
In addition to shedding light on the internalisation process, this analysis also contributes
to the crowding literature more broadly. While there have been numerous empirical studies
undertaken, as noted by Rode et al. (2015), many suffer from severe methodological
problems, especially in accurately measuring the base case of internalised motivation and
separating the effect of the policy from other factors. We consider that our methodological
approach deals with many of shortcomings that have been identified. We also provide a rare
case of crowding in, especially in relation to the case where a charge has been imposed to
deter behaviour13.
9. CONCLUSION
This article explored the process of internalisation of the law observing the impact of the
plastic bags charge in England. Focusing on primary data and on a difference-in-difference
estimator, this article observes that the introduction of a charge on plastic bags used in large
English retailers. Results indicate that the charge reduced new plastic bags use by 1.5 bags
per shop after 7 weeks. Importantly, this change was driven by a change in internalised
attitudes toward plastic bag use in addition to any pure price effects. In particular, the policy
made consumers more accepting of the government’s role in regulating plastic bag use,
increased intrinsic and moral motivation in reducing their use, and facilitated the coordination
problem between consumers.
13
As found by Rode et. al. (2015), such charges are normally associated with crowding out
effects generated by control aversion, frustration, reduced warm glow, reduced image
motivation, release from moral responsibility and frame shifting
APPENDIX A: QUESTIONNAIRE
Demographics
1) Where do you live in the UK? (England; Northern Ireland; Scotland; Wales)
2) What is your age?
3) What is your yearly income (before taxes)? (Below £ 15,000; £ 15,000-£ 20,000; £
20,001-£ 30,000; £ 30,001-£ 40,000; £ 40,001-£ 50,000; £ 60,001-£ 60,000; £
50,001-£ 70,000; Above £ 70,000; I prefer not to say)
4) What gender do you identify with? (Male; Female)
5) What is the highest level of education you have attained? (Basic education;
Secondary education; University degree – undergraduate level (e.g. BSc, BA);
University degree – Master level or equivalent (e.g. MSc, MA, MRes); University
degree – Doctoral level (PhD); Other - please specify)
Behavioural questions
6) In which of the following retailers have you been grocery shopping in the last
month (choose as many as you need)? (Asda; Aldi; Co-operative; Iceland; Marks &
Spencer; Morrisons; Lidl; Sainsbury's; Tesco; Waitrose; Small local retailer; Other
– please specify)
7) Think about the last time you used carrier bags in your grocery shopping trip (these
can be your own bags, as well as new ones from the store). How many bags did you
use? (if above 15, please indicate 15)
8) Of the bags you used, how many new carrier bags did you get from the store? (if
above 15, please indicate 15)
9) Do you think it is right for the government to make retailers charge customers for
plastic carrier bags? (Yes; No)
10) What do you think a fair charge for plastic carrier bag should be (in pence)?
Attitudes
On a 100-point scale (0 = “I strongly disagree”; 100 = “I strongly agree”), what is your level
of agreement with the following statement? (randomised order)
11) Minimising the number of plastic carrier bags when I shop for groceries is
important to me, regardless of any benefit or inconvenience that may result.
12) I feel good when I don’t use new plastic carrier bags when I shop for groceries
because it helps the environment.
13) Charging for carrier bags and giving the profits to a good cause (e.g. donating to a
charity) will ensure that plastic carrier bag use is no longer a problem for our
society.
14) The government should not interfere by requiring retailers to charge for plastic
carrier bags.
15) Plastic carrier bags are currently overused
16) Other customers will continue using plastic carrier bags even if I stop using them.
17) Shoppers have a moral obligation to minimise the use of plastic carrier bags.
Complementary/Substitute behaviours
18) Thinking about the past 7 days, how much money did you donate to a good cause,
in pounds? (e.g. donating to a charity)
19) Thinking about today, did you have a reusable carrier bag with you? (Yes; No)
20) Thinking about your last grocery shopping trip, did you buy any long-life/reusable
bags? (Yes; No)
21) Finally, thinking about your last grocery shopping trip again, did you buy plastic
bin liners? (Yes; No)
BIBLIOGRAPHY
BERTRAND, M., DUFLO, E. & MULLAINATHAN, S. 2004. How Much Should We Trust
Differences-in-Differences Estimates? The Quarterly Journal of Economics, 119,
249-275.
BOWLES, S. 2008. Policies Designed for Self-Interested Citizens May Undermine "The
Moral Sentiments": Evidence from Economic Experiments. Science, 320, 1605-1609.
BROWNING, M., CROSSLEY, T. F. & WEBER, G. 2003. Asking consumption questions in
general purpose surveys*. The Economic Journal, 113, F540-F567.
CARDENAS, J. C., STRANLUND, J. & WILLIS, C. 2000. Local Environmental Control
and Institutional Crowding-Out. World Development, 28, 1719-1733.
COMMONS, H. O. 2014. Plastic Bags. . In: COMMITTEE, H. O. C. E. A. (ed.). London.
COOTER, R. 1998. Expressive Law And Economics. The Journal of Legal Studies, 27, 585607.
ETZIONI, A. 2000. Social norms: Internalization, persuasion, and history. Law and Society
Review, 157-178.
FEUNEKES, G. I. J., VAN 'T VEER, P., VAN STAVEREN, W. A. & KOK, F. J. 1999.
Alcohol Intake Assessment: The Sober Facts. American Journal of Epidemiology,
150, 105-112.
FREY, B. S. 1992. Pricing and regulating affect environmental ethics. Environmental and
Resource Economics, 2, 399-414.
FREY, B. S. & STUTZER, A. 2008. Environmental morale and motivation. In: LEWIS, A.
(ed.) The Cambridge Handbook of Psychology and Economic Behaviour. Cambridge:
Cambridge University Press.
GNEEZY, U., MEIER, S. & REY-BIEL, P. 2011. When and Why Incentives (Don't) Work to
Modify Behavior. The Journal of Economic Perspectives, 25, 191-209.
GNEEZY, U. & RUSTICHINI, A. 2000. A Fine is a Price. The Journal of Legal Studies, 29,
1-17.
GOESCHL, T. & PERINO, G. 2012. Instrument choice and motivation: Evidence from a
climate change experiment. Environmental and Resource Economics, 52, 195-212.
GOVERNMENT, U. 2015. Carrier bags: why there's a charge.
Department
for
Environment
Food
and
Rural
[Online]. London:
Affairs.
Available:
https://www.gov.uk/government/publications/single-use-plastic-carrier-bags-whywere-introducing-the-charge/carrier-bags-why-theres-a-5p-charge
[Accessed
15
January 2016].
HART, H. L. A. 1997. The Concept of Law, Oxford, Oxford University Press.
IMBENS, G. W. & WOOLDRIDGE, J. M. 2009. Recent Developments in the Econometrics
of Program Evaluation. Journal of Economic Literature, 47, 5-86.
MCADAMS, R. H. 1997. The Origin, Development, and Regulation of Norms. Michigan
Law Review, 96, 338-433.
RODE, J., GÓMEZ-BAGGETHUN, E. & KRAUSE, T. 2015. Motivation crowding by
economic incentives in conservation policy: A review of the empirical evidence.
Ecological Economics, 117, 270-282.
SCHWARZ, N. 1999. Self-reports: How the questions shape the answers. American
Psychologist, 54, 93-105.
SCHWARZ, N. & OYSERMAN, D. 2001. Asking Questions About Behavior: Cognition,
Communication, and Questionnaire Construction. American Journal of Evaluation,
22, 127-160.
TITMUSS, R. M. 1971. The gift relationship: from human blood to social policy, New York,
Pantheon Books.
WOOLDRIDGE, J. M. 2003. Cluster-Sample Methods in Applied Econometrics. The
American Economic Review, 93, 133-138.
WRAP. 2015. WRAP announces new carrier bag figures.
http://www.wrap.org.uk/2015_carrier_bag_figures.
[Online]. Available: